The Natural Basis of Morality

Dr Donald Cameron, on 7 May 2002

Summary

Dr Cameron argues that three self-evident axioms provide a basis for a value premise that is provably correct. These axioms lead to the Evolutionary Value Principle — to prefer those actions that will increase the frequency of our gene alleles in the population. He then explains why there is altruism and suggests there are three types — kin, reciprocal and display. Evolution takes time to work so that modern harmful `instincts', e.g. to remain childless, have not yet been evolved out of our systems, but it can provide a method of solving current problems in politics, bioethics, animal rights, warfare, art, sport and attitudes to death. Evolution creates and is the only source of information that supplies order and prevents randomness in the arrangement of molecules in living things.

********

 

Victor Suchar, when arranging this talk, said "please be controversial, it makes things more entertaining" and it would seem impolite not to accept his advice. So I propose this evening to set out an answer that has eluded philosophers for millennia. I will give an objective definition of the meaning and purpose of life and a basis to answer every possible question of ethics. I will also offer a proof of this, which will be as sound as the theorems of Euclid. I admit that this sounds a tiny bit ambitious, but I would ask you to pay very close attention, so that you can end the evening enriched by what you have learned - or, perhaps more likely, be able to shoot down my argument in flames at question time!

 

Statements about value, purpose or morality are fundamentally different from statements about fact. They are a different kind of information. Fact information is a description of an external reality, whereas value information corresponds to nothing external. Value information is rather something that is contained within the mind of the decision-maker. It does not describe what the outside world is like; it describes how the decision-maker would like it to be.

Even in a decision-making machine, the distinction between fact information and value information is clear. A guided missile, for example, takes in external information about its position and that of the target, but its makers have built-in its objective function - in this case to minimise the distance between the target and its point of impact. This must be internal information, which it cannot get from the environment.

The first formal recognition of this truth is attributed to Hume in the late 1700's and is today often encapsulated in the saying that you cannot deduce an "ought" from an "is". Some modern philosophers deny that we must draw a sharp line between "is" and "ought". I have yet to understand an argument that could convince me that they are right.

But it is important to realise that this distinction between facts and values is a first-person thing. It is only a clear concept if it is applied to an individual decision-maker. A value that any of you might hold would, in my brain, be a fact - it would be information about my external world. I would not be recruited to it automatically unless I have an internal value that gives me a reason to be so. It is very important to be clear about this. Religious people say that the source of value information is the will of God, but the will of God, to me (if God existed) would be a fact. Why should I be interested in it? It would not be my value, unless I have an internal value to respect God's will, or perhaps to fear his punishment.

To clarify this, let us suppose that someone says that he believes in his values because they are the will of his pet rabbit. This is obviously ridiculous, but why do we feel that God is less so? The answer is that we have innate preconceptions of how we should value the will of someone who is in a position of powerful authority.

We can now extend Hume's Law a little. I cannot derive a value from premises that are composed only of facts, and I cannot derive a value from the values of others because these constitute only facts in relation to my brain. This principle is hard to recognise because we have internal values to conform to the norms of our society and to harmonise our values, where we can, with our friends and neighbours. This is a complex instinct, which includes the skill of deciding when to trust and when to be suspicious.

It follows from Hume, that we can never obtain our values from others. Others can suggest values, but we will only accept them if we have an internal will to do so. Of course we can be persuaded (and even deceived) into accepting the values of others, but this is always by becoming convinced that they meet the aims of our pre-existing values. That conviction may arise by reason, or by instinctive acceptance, and it may be correct or in error. It is the job of philosophy to find ways of protecting us from error.

Although we must not forget that the only ultimate source of our values must be internal, we can do a lot of useful reasoning about our values, both what is worthwhile for ourselves and our ethical values towards others. Fact information can be relevant to work out a value, but no deduction of value can be made without at least one premise that is a value statement.

A typical example of this is a situation where I see a man in difficulty in the water and I have to decide whether I ought to throw him a lifebelt, which happens to be to hand. The fact that he is in difficulty is not alone a proof that I should throw it, I must also hold a value premise, for example, that I ought to help those whose lives are in danger. But although facts alone cannot lead to a value conclusion, taken together with a value premise, they can completely alter it. Suppose, for example, I discover that the lifebelt is a solid lead replica.

Before considering ethics, our purpose in relation to others, I would like to look more deeply into what is the purpose of anything at all. I was discussing this with Professor Robin Downie of Glasgow University Philosophy department and he explained that enquiries into the meaning and purpose of life had not been much written about or discussed during the last century or more. His implication was that this was a good reason why we need not study the subject, which is quite out of fashion. To ask what is the purpose or meaning of life has become something of a joke. By far the most productive contributions have come from Douglas Adams and the Monty Python team. Yet it seems odd that moral philosophers have tried to construct solutions to the multi-person problem before tackling the very much simpler, single-person problem. As we have just discussed, Hume's Law is only clearly defined when applied to the single-person case.

Lord Kelvin famously said "When you cannot express it in numbers, your knowledge is of a meagre and unsatisfactory kind; it may be the beginning of knowledge, but you have scarcely in your thoughts advanced to the stage of science". His dictum is not true in every case, but it certainly is, in the subject that we are studying tonight. We can understand our problem better by turning to the methods of mathematical decision theory.

When we intend to solve any problem mathematically, we are forced to clarify our ideas and, in this case, it brings the two types of information into sharp relief. Fact information is used to create a mathematical model of that part of the external world that is of interest. The model can then be used to predict the possible future states, which could result from actions that are open to our decision. The value information is quite different: it consists of an "objective function" whose parameters are the states of the system, but whose result is a one-dimensional quantity, a value, which must be maximised or minimised.

Operational Research is the practice of applying mathematical decision theory to military, business or other management problems. One of its most notorious difficulties is to define the client's objective function. A business manager may want to maximise profits, maximise benefits to the employees, give the best possible service to customers, achieve complete safety and be caring of the environment. Sometimes managers or politicians like to leave the weighting that must be given to these competing objectives undefined, so as to deceive their hearers. Yet no optimum decision can be calculated until this is done. A single, one-dimensional objective function must be defined for any precise analysis of decision.

In business applications this is sometimes achieved by reducing everything to a money quantity, sometimes stretching a point to do so. Questions, such as how much money we are willing to spend to save one person's life, are difficult, but it is useful to answer them. If they are not answered, and decisions are made in the old fashioned way, these can be reverse-engineered to find the values that were effectively chosen. That may reveal inconsistencies - for example, we can expect to find that the value given to a human life is greater where extensive publicity is likely!

As individuals too, we have many conflicting goals, but to function as a decision-maker we must specify the weighting between them. At times this is an unconscious process. Obviously money is not the bottom line - that could only be a sub-objective. Even if we do not specify the weighting of our sub-objectives consciously, our actions after the event will define what our choice effectively was.

Our problem, then, is to discover a provable scientific basis to define our own ultimate objective function, yet we have just shown, in our discussion of Hume's Law, that it is impossible to find that anywhere other than within ourselves. How can we hope to create a science of values? We can apply reason to our values, but any deduction of a value statement must have at least one premise that is a value statement and that can only come from within ourselves. As Hume said, reason can only be the servant of the passions.

Now this situation has been used by most philosophers as a licence to grab fully formed, complex value statements out of their emotions, which have been conditioned by the fashionable culture of their times. As their reasoning leads them into distasteful contradictions, they feel free to patch things up by using phrases like "but most right-thinking people would find that repugnant", thereby slipping in a further premise. An example might be to start with the premise "it is wrong to lie", but we then ask if it remains wrong to lie if a murderous madman bursts into your house and asks where your children are sleeping. Then there is talk of exceptions and "fuzzy edges". I have even heard it said that it is a good thing that there are fuzzy edges! Perhaps we could call that statement meta-fuzziness!

Some people, mistakenly in my view, claim that there are intrinsic values, which exist independently of the people holding them. We have to ask in what material sense could they exist? Where is the information stored? There are no tablets of stone or writings on cosmic spheres. And if the information was somewhere - carved on a big block of stone on a mountain top for example - why should I be convinced to obey it? We have the same problem that we had before. These values would be external facts; we will only adopt them, if we can be persuaded that they will serve our internal values and instincts.

It has been said that intrinsic values are those held by all right-thinking people, but we can see what is coming next. Any counterexample can now be explained by branding some of the people as wrong-thinking. The values that are being described as intrinsic are simply those that are either built in to the majority of people, or accepted by the majority. The information does exist, but only in the brains of those who hold the values.

The idea that there "is" some disembodied moral imperative becomes absurd when closely examined, although it may give a practical rule that works. An uncritical acceptance of it may well result in an individual doing the "right" thing, but as philosophers we really ought to understand its mechanism. I will return to that in a few moments.

Some, for example, might talk of achieving the greatest happiness of the greatest number. Consider a woman who reads in a newspaper that 1000 children have been killed in a Venezuelan earthquake. Now consider an alternative scene in which a policeman arrives at her door and tells her that her own child has been killed in a road accident. Which do you suppose would cause her the most grief? Our philosopher will tell us that, although her concern for her own child is understandable, at a fundamental level she ought to feel more for the 1000 children on the other side of the world than for her own single child. My answer to the philosopher would be: "No sir, it is not the woman's instincts that are imperfect - it is your philosophy that is hopelessly wrong!"

It is absurd to talk of maximising happiness as if it were a quantity; it is even more absurd to suppose that it is a one-dimensional and additive quantity. It is obvious that happiness is not a quantity, it is a mechanism. The whole procedure is messy and not worthy of the name of science. After thousands of years of development, philosophy's reputation for uselessness is sadly deserved. The need to define a one-dimensional objective function is not recognised and certainly not achieved. Happiness of the greatest number fails on several counts.

I do not wish to show too much disrespect for the philosophers of the past. Their achievements were great with the information that they had available, but prior to 1859 they were doomed to failure. Despite this I am awed by the clarity which Hume was able to produce in the late 1700's.

But today, equipped with the work of our predecessors, we do not need exceptional talent to do better than that. If we are to construct a proven system of values, we cannot escape the need for at least one starting value premise, but can we do it in a more orderly way. If we could start with basic premises of value, self-evident axioms, which are simple, beyond dispute and independent of the fashions of our time, we could build a science of values that is, in some sense, provably correct. That is how Euclidean geometry, the very model of deductive perfection, is built.

The problem is how to choose the starting axioms. They must contain enough information to construct a complete system of values, yet not so much that they create contradictions or assert that which must be deduced. (The statement "it is wrong to lie" had these kinds of problems.) The choice of axioms is an expression of the individual and, certainly, more than one axiom set is possible. One choice, which I would like to explore tonight, is the following:

(a) to wish not to hold contradictory values. (few would dispute this one)

(b) to reject nihilism. (the idea that nothing matters at all) (this one is more a matter of my choice, but in practice I believe it will be the choice of most human beings. That doesn't make it right of course. An artificial intelligence machine might reject it - it could not be logically faulted for doing so.)

(c) to wish one's values not to be a result of random accidental events, but to have a source of information. (This one is less self-evident, but overlaps partly with the other two. If we allow our values to come from random happenings the occurrence of contradiction is likely. Also, the acceptance of values that arise from random occurrences seems close to saying that no values matter at all.)

Now the really interesting thing about this particular set of axioms is that it is possible, taking them together with fact information, to work out a complete science of values which answers every possible question of value and ethics. No other value premise has to be introduced. There are no exceptions and no fuzzy edges. And the result comes very close to the values that arise from our intuitive common sense.

If this project is indeed possible, as I am alleging, then it might be something of a contribution to the problem. Such an ambitious claim must be carefully scrutinised, however. Let us see if it can really be demonstrated:

We have talked about value-statements and their derivation, as they must be, from premises of value, but, for the moment, I would like to stop doing that and to discuss only facts.

Millions of years ago, following the big bang, the universe was a mass of non-living matter and values could have no meaning. There were no decision-making devices or organisms and so there could be no objective functions. Concepts such as right and wrong, desirable and undesirable did not exist. After much time, replicating molecules started to generate more complex structures, including life forms that processed information. Even the most primitive of these had an objective function. The animal's decisions were directed more towards some real-world outcomes than others: obviously they will prefer to eat and not be eaten, to survive and to reproduce. Random events followed by selection can produce non-random results - the process can be a source of information. That is why we are complex, information-rich organisms today. Random events alone can never produce a non-random result.

The only source of the very large quantity of non-random information that comprises the human body and mind, including its values and the human instinct to build an ethical culture, is the force of natural selection. The fact of evolution and, in particular, the modern analyses of the evolution of altruism and social behaviour are essential to understand any philosophy of values. It is astonishing that so many investigators of ethics have felt able to ignore them. (I have a quick way of evaluating any book on ethics. Look for "Darwin" or "evolution" in the index. If both are absent, discard it at once!)

Some may say that there could be other sources of this information. It remains fashionable to believe that values could be an emergent result of culture. "Emergent" is such a meaningless and ignorance-hiding word! Culture is information that is passed between human beings. Some of it is about observed facts and some about values. The idea that values could originate in the process of communication, rather than in the human brains, can be seen to be nonsense as soon as it is closely inspected. This is an illusion, which I am tempted to call the culturalistic fallacy (in contrast to the naturalistic fallacy, which is the attempt to deduce values from facts alone). Of course a more highly developed culture often results in a more altruistic level of morality, but that is a mechanism which is a consequence of people's values, not a source of them. We will study the reason for that in a few moments.

When we examine the intuitive values of real people we find that these conform, in astonishing detail, to the values that would be produced by natural selection - this is not entirely surprising, since natural selection has indeed produced them! Modern intuitive values differ from the evolutionary optimum at times, but these are mostly in relation to modern choices, where our species has not yet had time to adapt to a new (and often man-made) environment. Modern values such as enjoying cigarettes, or remaining childless by choice, are accidental interactions between our evolved instincts and modern circumstances.

Axiom (c) says that we must find an information source for our values, but does not specify what it is. Perhaps several are available and we will still have to exercise a value choice between them? In fact this is not so, because there is only one: the force of natural selection. This leaves our individual purpose as being to prefer those actions that will increase the frequency of our gene alleles in the population. For ease of reference I will call this the Evolutionary Value Principle(EVP). Should this, then, become our guide?

Philosophers will be quick to point out that we can accept the fact that our brains, nervous systems, feelings of value and even our instincts to create systems of morality, may have been created from randomly distributed molecules by natural selection, but that is simply a matter of fact. It carries no compulsion on us to continue to promote the values that we have inherited. To think otherwise would breach the is/ought rule; it is the classic naturalistic fallacy.

And of course they are right. Our values have to come from a choice within ourselves, but what our reasoning has shown is that we can only reject the evolutionary value principle, if we reject the choice of axioms with which I started, or to find a flaw in the deductive steps from them. To reject the axioms is to admit to acceptance of values whose source is random accident - or to no values at all - or even to values which contradict each other. The person who says that no values matter at all is logically faultless - his position simply doesn't appeal to me, because I am an evolved animal. But if there is a value system which says information drawn from random events can be admitted, or that self-contradiction is acceptable, let no one try to convince me that reason could commend them as important, or that any resulting moral duty is binding.

Every part of human reasoning depends on the innate values, logical rules and facts that have been programmed into our brains by evolution. For example our brains expect space to have three dimensions, not two or four, and we have the basic rules of logic (for example "if A -then B" taken together with "A is true" must mean that "B is true"). These building blocks of mental function seem to be innate. Concepts of value such as fairness, as well as some very specific values such as caution near precipitous places or snakes, seem to be in our brains without learning. Some of these seem very accurate while others are poorly developed. We do not seem to be good at the assessment of small probabilities - without that imperfection, lottery tickets would never be sold.

Evolution has equipped our minds with many preconceptions of value and, from the way that natural selection works, we can see that it will be a patchwork of different values sometimes overlapping, sometimes duplicating, but always serving the same ultimate end which is the EVP. An example of overlapping innate values might be the wish to stay away from cliff edges and the wish not to die.

The axioms, which I have pulled out of my instinctive feelings, are only a very small subset of these innate values. I feel instinctively sure that they are true, but I cannot prove them. This seems to be a weakness, but it is one that is shared with every result of logic, mathematics or science. But the fact that they are sufficient to derive most of the other innate values shows that the whole procedure is not very sensitive to the choice of this subset. The only values which humans typically show which are not aligned with the EVP are those arising from modern circumstances to which natural selection has not had time to react.

Some may be dismayed by this analysis which at first sight seems to negate all morality and altruism. George Bernard Shaw said of Darwin's Theory: "It seems simple because you do not realise all that it involves. But when its whole significance dawns on you, your heart sinks into a heap of sand within you. There is a hideous fatalism about it, a ghastly and damnable reduction of beauty and intelligence, of strength and purpose, of honour and aspiration." He would not have liked the use of the theory of evolution that I am making tonight.

But if the news is bad, so what? Many people use the argument "Oh, that would be very unpleasant, therefore it could not possibly be true". Sadly the world is full of counterexamples. They show that, although denial can offer comfort, it is useless as a way to find the truth. Richard Dawkins is right in his answer to those who complain that they do not like what evidence shows. His answer was "All I can say is, that's just tough. We have to face up to the truth". (Incidentally I came across this quote accidentally on the web page of the American "Answers in Genesis Ministries" who wanted to show how heartless an evolutionist/atheist was.)

Those of you, who have listened this far without recoiling in horror, will perhaps be saying "Does this mean that there is nothing but selfishness in the world and that all our hopes for progress are dashed?" Certainly it seems a puzzle. If it is true that the best way to propagate our genes is to be selfish, the nature-red-in-tooth-and-claw argument, why do we see so much altruism around us? Why has it survived the relentless force of natural selection? What countervailing force could have defended it from extinction? Why do we have instincts to reject a life of total selfishness? In fact, altruism is here because it has evolved and, for exactly the same reason, it is commended by the Evolutionary Value Principle for the future. Perhaps things are not so bad after all. Let us examine in more detail by what mechanisms altruism could evolve.

The first type of altruism is kin altruism. Your own children share 50% of your genes and other relatives varying proportions. Obviously evolution would select for the instinct to help one's own offspring. By promoting their survival and reproduction you will be promoting the copying of your own genes. Thus unselfishness, at the level of the individual, can be explained by selfishness at the level of the gene. This was the idea explained in Dawkins's The Selfish Gene, which some silly people misunderstood to mean a genetic justification for selfishness. There is no doubt that pure, selfless altruism towards kin has evolved and equally is commended by the EVP.

The proportion of genes shared by relatives peters out very quickly. A sibling or child has 50%, a cousin 12.5% and a person who shares one set of great-grandparents 3.1%. Yet we see altruism extending far beyond that. This is explained by reciprocal altruism. Human beings have much to gain by cooperation, and mutual trust. Individuals who fail to take part in this will be at a selective disadvantage.

A much discussed situation is the "prisoner's dilemma", where two people must decide to either "cooperate" or "defect", without knowledge of how the other player will decide. This game encapsulates, in its simplest form, the basic essence of the cooperation problem. A version of the game might give rewards as follows: if both cooperate then both will receive seven dollars; if both defect they will each gain three dollars; if one cooperates and the other defects, the cooperator will receive nothing while the defector will receive ten dollars.

The dilemma is clear. The total benefit is greatest if both cooperate and if both defect then both are worse off. But if I cooperate and the other player defects, then I am a sucker and lose badly. Even if the other player cooperates, it would still pay me to defect. How can cooperation be established? In a game of a single play, it probably can't. In repeated plays, however, it may be different.

Computer simulations have been run by Robert Axelrod, in which he allowed "individuals", each consisting of a small scrap of computer program, to play against each other. After each round of play, they died and reproduced according to how many points they had won. At first, the aggressive defectors multiplied at the expense of the suckers, but, after a time, their numbers declined as their prey became extinct. The long-term victor was a routine called TIT FOR TAT, which began by cooperating, but punished its opponent for defecting. It was concluded that a strategy which would give selective advantage must be "nice" (always cooperating first), "provocable" (not leaving defection unpunished), "forgiving" (quickly re-establishing cooperation after a defection) and "clear" (having rules of behaviour which its partner can identify).

This model of human interaction is grossly oversimplified, but it has some interest and goes some way to explaining how cooperation could have evolved. In the real world, cooperation may be between two people, but is more often a multi-person game. A person may be altruistic to an inexplicable recipient, who cannot reciprocate, so as to consciously or unconsciously establish himself as a good cooperator in the eyes of a third person. This type of altruism I have named "display altruism". The organisers of charity social events understand this mechanism perfectly.

Many people hold that certain principles are simply right, without question of personal advantage, and are held to be so by all right-thinking people. By this they mean that there is a moral imperative which somehow "exists". They might, for example, say, "I think it is fundamentally wrong to kill another human being". Equipped with our understanding of the nature of human altruism, we can see what is really happening. What such a person is really doing, without being conscious of the fact, is taking part in a multi-person cooperation, in which a society's members accept and enforce the upheld rule. By doing so, he will gain more by not being killed than he will lose by foregoing the convenience of killing others. There is probably a measure of display altruism taking place also.

These culturally transmitted values are better understood as being multi-person, partly unconscious, reciprocal cooperations whose duration may be longer than the lifetime of any of the individual participants. The participants, moreover, are playing the right moves because their instincts to do so have been selected - not because they really understand the mechanism. The alternative description of cultural values as an emergent property of cultural development does not entirely contradict this, but lacks any explanatory value. Here is the origin of the idea that values could come out of thin air during the development of culture. The more civilised a society is, the more it has invested in building up this cooperation software, complete with its anti-defection mechanisms. That increases the gain to all of its members through multi-person, reciprocal altruism.

Evolution, in forming these instincts, had only to create behaviour to play the right moves; there would be no selection pressure to create conscious understanding, unless that were the best way to achieve the result. A brain that believed in a moral imperative or an intrinsic value would do what was required and its genes would be positively selected. We can begin to understand why it so difficult for animals like ourselves, the products of natural selection, to stand on their hind legs and become moral philosophers!

And we can now see why the Evolutionary Value Principle is not a simple recipe for heartless selfishness. In fact, it supports all of the altruism that we see around us. A "better" culture of values is more likely to appear, if we understand that we are building together a reciprocal cooperation than if we waste our time on impractical utopias. But this principle does not exclude all competition either: only the naive could have expected that.

So what are the practical applications of these ideas? Unlike most philosophical theories, the Evolutionary Value Principle actually has some that we can go straight out and use. Firstly, reproduction is not just the main thing: it is almost the only thing. Childlessness by choice is a disaster akin to suicide or murdering our own children. The only reason we view these choices with more horror than life-long birth control is that the new choice has not been around long enough for our instincts to have adapted.

Mr Eric Stockton, from Orkney, wrote to me to say that he agreed with most of my theoretical analysis, but found the conclusion that one should have more children to be, in his words, "just plain dotty". My reply did not include any philosophy of value - just an uncomplicated piece of arithmetic. At the present rate of reproduction, the indigenous people of Europe will have declined to 3% of their present number after 200 years. It seems a sad end to their long history. I would emphasise that I do not blame others for this disaster. The populations who are moving into our countries to replace us cannot be blamed and may suffer in their turn from the same disease. The problem is our instincts, which, in their reaction to very recent modern conditions, have departed from the EVP. That is what is really "dotty".

Some commentators have speculated that evolution, for humans in any case, has come to an end. It is certainly true that we have succeeded in controlling many of the dangers that beset people in a more primitive state of technology, but natural selection is working upon us faster than ever. The instinct to remain childless for social convenience will certainly be weeded out in a few generations; we cannot tell if it will be quick enough to save our descendants from the arithmetic that I have just described. The effect of this instinct on the gene pool is similar to an instinct to jump out of high windows. Instincts to enjoy motor cycling, bungee jumping and other modern activities are probably coming under selection pressure too.

The Evolutionary Value Principle also commends the individual to be a good reciprocator, to earn a good name for integrity, to uphold the norms of a civilised culture, to pursue learning and to create wealth. But all of these good things are servants to the recommendation to reproduce.

In addition, the EVP provides us with an algorithm by which we can solve every problem of values in politics, bioethics, animal rights, warfare, art, sport and attitudes to death, without fuzzy edges and without grabbing value statements out of the air.

For example, consider the couple who had an embryo selected so as to be able to provide transplant material (harmlessly) to its older sister who would otherwise die. Many of the philosophers, luminaries, priests and other tele-pundits are out in force to declare this unethical, yet we can calculate the answer. Clearly it is in the parents' interest to have their first child survive and to have subsequent children free from the disease. It is even in the interest that we might calculate for each child. The achievement of the doctors seems to be an unqualified good for their patients. The alternative of allowing their daughter to die and choosing between having no further children or running the risk of the disease again is clearly inferior.

Some commentators have said that we should oppose such embryo selection because it might be the thin end of the wedge. If we allow it, they say, people could use it to produce "designer" babies. They could select the sex of their child or arrange for it to have blue eyes. This is a really convincing argument. Yes, it is important that we condemn this child to pain, suffering and an early death. It is right that we permit the grief of the parents. We must be prepared to make that sacrifice (which gives no pain to ourselves, you understand) so that the terrible evil of a few more blue-eyed people about in the future can be prevented.

Such absurd moral reasoning illustrates the need for a clearer moral philosophy. In traditional situations people tend to reach reasonable moral decisions using the innate feelings that evolution has provided them with. It is when we encounter situations that have no precedent in our evolutionary past, such as those raised by new medical possibilities, that we are most likely to make bad judgements. That is when we need the general understanding and power of extrapolation given by the EVP.

Most people today agree that Darwin's Theory of Evolution is true. Even the Pope agrees it is true. Yet so many of them continue live with the contradiction of accepting the theory, but ignoring its clearest consequences, perhaps vaguely accepting some kind of duality, in which the medieval soul inhabits a body descended from an ape. Some think that evolution happened, but was steered by God. There is no evidence for this steering. Some believe that we evolved by natural selection but at some point we suddenly acquired souls to become people instead of animals. There is not the slightest evidence for that either. We might imagine what it would be like to be the first generation to have souls, raised by parents who were mindless automata. I suppose that is not too different from the opinion of my own adolescent children!

Some years ago, the British comedian Kenny Everett gave an illustration of the chemical composition of the human body. He obtained the appropriate chemicals and mixed them in a tub, but no human being emerged from the sticky mess. The reason was that one ingredient was missing: that was information. The arrangement of the atoms that make up any living thing is extremely non-random. As Richard Dawkins has said, there are many more arrangements of the atoms that will die than those that will live.

The existence of order, the non-random arrangement of living things, has always required explanation. Paley, in his Natural Theology of 1802 pointed out that a stone on the ground would not be remarkable, but a watch discovered in the same place would point to the existence of a watchmaker. From this he argued that the very non-random nature of life proved that there must be a God. Of course this evidence did not prove the existence of a god with all the attributes written in the scriptures, but it did prove the existence of an information source, a source of the evident non-randomness. God's existence was a very good explanation, which was unchallenged until Darwin discovered that it was not the only one.

Evolution does not involve blind chance - the complex forms of life could never have evolved by chance. It is something like the idea that a junkyard hit by a hurricane might assemble a perfect Boeing 747. While perhaps not utterly impossible, it is so unlikely as to be not worth considering. The missing ingredient is the information, the order, the non-randomness; for living things, that comes from the mechanism of natural selection. Natural selection creates information by selectively preserving a non-random subset of the mutations that randomly occur.

Creation science is a movement, which is active, particularly in the United States, to oppose the idea of evolution, although the use of the word science is a bit odd. Science studies the evidence to try to discover the answer. Creation science is committed to the answer and tries to discover the evidence.

Creationists often claim that evolution contradicts the second law of thermodynamics. In its direct application to the distribution of heat energy, this is clearly false. The sun has supplied enough high-grade energy throughout evolutionary history. There is an analogy, however, between information and entropy, which makes some sense. Just as the entropy of the universe, or any closed part of it, can only increase, information flowing in a closed channel can only degrade. If the signal can not be added to, it will ultimately decay into noise. The similarity is closer than this analogy suggests, with a close correspondence between the mathematical formulations of thermodynamics and information theory.

But the second law of thermodynamics is not absolute. It is simply a statement of probability, but with such a large number of atoms in the universe, the statistical average inevitably prevails. Yet, in the 19th century, James Clerk Maxwell presented a counterexample. He imagined that two chambers of gas at the same temperature might be connected by a small aperture. At the aperture sits a small demon with a table-tennis bat. As molecules approach the aperture from chamber A, he bounces back all the fast moving molecules, but allows the slow ones to pass. Molecules approaching from chamber B are bounced back if they are slow, but allowed to pass into chamber A if they are fast. In time the gas in chamber A will be at a higher temperature than that in chamber B without any external source of energy and the second law will be broken. The demon has added no energy - only information. We confidently use the second law today because Maxwell's demon does not exist. But if some future nano-technology could ever produce a machine to work as a Maxwell's demon, we would have a solution to the world's energy problems at a stroke.

But when we come to the information content of living things, the situation is different. The analogue of Maxwell's demon is alive and well and it is called natural selection. This force has been present for as long as replicators have existed, steadily producing non-randomness by selecting some and discarding others. This is why the creationist appeal to the second law is as false in its information theory analogy as it is in thermodynamics.

All living things are highly ordered or non-random. They contain information. Our minds are the function of our nerve cells; our nerve cells have the same source as our muscle cells, bone cells and blood cells. They are all the products of evolution. All of these have been formed by many random events that have then been sifted by natural selection. Natural selection is the only source of information in our beings. We are made of very ordinary atoms and many of these atoms are probably not the same ones that were in our bodies ten years ago. But the essence of a person is the information content embodied in the arrangement of the atoms. This information has come from natural selection interacting together with random events, but the only non-random source is the natural selection. Natural selection has created us. Natural selection is essentially what we are. What a monumental waste of time it has been for philosophers to try to understand our purpose and ethics while ignoring it. The principle of non-randomness is central to our understanding of these things.

When asking people to accept a different approach to a problem there is always a problem if they are asked to change a belief which they have had no reason to doubt. The difficulties are magnified if they must be asked to change a number of such interlocking beliefs before the new idea can be understood. Of course, many of the concepts on which my argument depends are not original. I have drawn on biology, operational research, game theory, information theory, computing and mathematics. These are well established, but are perhaps too little considered in philosophy.

Let us review the main concepts that we need:

1) Values are internal, one-dimensional functions coded within decision-makers.

2) Acquired values can only be derived from internal values, not from facts alone. By extension to Hume's law, third-party values are facts and cannot alone be a premise for the deduction of first-person values.

3) The only information source for our internal values is natural selection.

4) The Evolutionary Value Principle is the only complete definition of value that has an information source and does not involve random interactions. The only other possibility is to adopt the position that no values matter at all. But to avoid that, we have required another axiom.

5) The axioms that we use - to reject values from an obviously random source, to reject contradiction and to reject nihilism - can only have come from our evolution. It is a consequence of Hume that we can only discover our true purpose, by using values which evolution has programmed into us, taken together with an understanding of the mechanism by which that came about.

6) We must avoid the culturalistic fallacy. The idea that values could have originated in the process of culture development is mistaken. What is really happening is that increasingly effective moral cooperations are being formed.

7) The superficial idea that the EVP could mean only selfishness is shown to be wrong by the recent understandings of the evolution of altruism. Morality is instead justified by the EVP in the form of kin, reciprocal, display altruism, and the building of civilisation.

8) The mistaken notion of a moral imperative is a misinterpretation of the instinct to engage in reciprocal cooperation. The mistaken notion of the utilitarian is a misinterpretation of our instinctive sense of fairness.

It is a big pill to swallow; it is very difficult to persuade anyone to absorb eight new concepts at one go. But the prize is enormous. It is an understanding of what our lives are about. Because it is derived on sound principles, it does not go wrong as soon as it is extrapolated; it does not go fuzzy at the edges. It can be extended to provide an answer to every question of value and ethics.

As I said at the outset, I would accept Victor Suchar's advice to be controversial. The other side of that coin is that I must expect you to produce a barrage of criticism. I am happy to receive that with goodwill and answer any questions that you may wish to put.

Although I would place no restrictions whatever on the questions that I will take, I would like to make one suggestion. I have, this evening, presented a deduction. To disprove a deduction you must dispute at least one of its premises, or point out a flaw in its logic. It is very much less interesting to point out that its conclusions are contrary to expectation or conventional wisdom.

So at this point, Mr Chairman, with your permission, I would like to invite questions.

Donald Cameron

A vigorous discussion followed but cannot be reported since transcripts of the lengthy and complex contributions are not available.

Editor