Consciousness and our Duty to Animals

Consciousness and Our Duty to Animals
Dr. Donald Cameron
BRLSI Member, Convenor Philosophy Group
1 May 2007

Consciousness
Before embarking on a discussion of our ethical attitudes to animals, I would like to consider the question of consciousness. Of course, we can question whether consciousness has anything to do with our problem, but there is at least a popular idea that it does. Some people would express discomfort when a chicken’s neck is wrung to prepare it for supper, yet would remain calm when some carrots are uprooted and chopped up for the same purpose. To take the matter even further, anyone who said that breaking up stones with a sledgehammer was cruel to the stones, would be thought ridiculous. The popular feeling seems to be that we have duties to naturally occurring things that can process information, but not to those that can’t.
There is much talk among philosophers about the “problem of consciousness” and I have read and listened to a great deal of disappointing material about it. I think the reason that there is such a problem finding the answer is that we have not thought enough about the question. I know that I am conscious, but I have no direct evidence that other people are, let alone animals. Yet I do have some reasons for believing that other people probably are conscious. They seem to be similar to me in many ways and are descended from the same evolutionary process. They appear and behave much as I do and they even declare themselves conscious.
Of course, it is difficult to be sure that they and I attach the same meaning to the word. Ultimately we can only learn the meaning of a word by association, or by its definition using other words. We can be reasonably sure that we use the same meaning of “ball” or “spoon” or “chair”, but no one has ever pointed to a consciousness.
Yet I think it can be defined in terms of other words. It means something about my internal thoughts and capability for action. It is not only that I am thinking or processing information, but that I know that I am thinking. Consciousness is something I lose when I go to sleep, but which partly returns in dreams. It is something to do with the quality of my brain function, but that is made up of several different attributes. What is clear it that it consists of information flows.
Perhaps a good way to clarify thought on this is to consider what it would take for a machine to be conscious. Computers at present do not achieve this, but given the rate of progress that they have shown in recent years, it would be rash to say that it is impossible that they ever will. Indeed, because we know that our brains are naturally constructed information processors that yield this property of consciousness, we can be sure that it is possible in principle.
Let us imagine a machine of the not-too-distant future with advanced technology. Let us suppose that our machine has the following characteristics – none of them impossible to imagine as an extension of what can be done today:
a) It contains information defining an objective – a description of preferred states of its external or internal world.
b) It has some useful information installed in its memory at manufacture and also has rules of logic and induction pre-programmed.
c) It takes in information from sensors, perhaps focusing light images, detecting sound waves, deriving information from touch, trace chemicals in the atmosphere or in solid and liquid substances. Humans do all of these, but our machine might also detect radio waves, infrared light, ultrasound or other sources of information.
d) Like humans, the machine will reduce the mass of sensory information into theory models of the outside world. The formation of theories or mental models, as has been explained elsewhere2, is no more than the distillation of a mass of data that contains redundancy, to produce a model of the world that can be used for prediction.
e) Our advanced machine has outputs. It can generate decisions, using this information, to maximise its objective (or at least to do better than random).
f) The machine also takes in information from internal sensors, both from its physical structure and from parts of its information processor. In particular, it keeps a model of aspects of its own thought process and can report that it is actively processing information.
g) It will have an information store or memory where observations, theories and past decisions can be kept for future use.
h) We will assume that our machine is capable of transmitting information about its observations, its theory models, its decisions, its objectives and its internal states to humans or other machines. Information so transmitted must be in some kind of code whose elements relate to each thought representation. Perhaps it can speak English.
i) And finally let us suppose that it contains in its objective-setting information, an overriding objective that it should not be switched off or destroyed. This objective has a value above all others so that the maximising mechanism will cause outputs that achieve its preservation, if there is the slightest possibility to do so.
Such a machine is not beyond the imaginings of computer engineers for the not-too-distant future. It would clearly follow its own objectives (they are built in). Of course, these objectives would come ultimately from the objectives of its constructors. If it detects that anyone or anything will try to destroy it, it will generate outputs to the limit of desperation to prevent that. Is such a machine conscious? It would certainly declare that it was, if we gave it a description. Would it be immoral to switch off or dismantle this machine?

The Sense of Self
But is consciousness sufficiently defined by this description of information flows? Some would say that there is something missing – the essential feeling of self, of what it is like to be me that is different from other people. Some understanding of this problem can be gained from the teleportation thought experiment.
Of course, teleportation of the Startrek style is unlikely ever to be possible, let alone safe, but let us suppose that it has been perfected. How might it work? Perhaps some way would be found to measure the type and position of every atom in your body and transmit this data to a distant site. At the receiving station, using a stock of raw materials, atoms would be assembled in exactly this pattern to manufacture a living person. The resulting product would be you according to every test that could ever be devised. Your nearest and dearest would recognise it as you. It would have all your memories and think it only right to be entitled to your bank account and to sleep with your spouse. It would possess all the consciousness of being you.
But back at the transmitting end, “you” are still there. The idea that you should now be dissolved and your atoms put into the raw material bins to be used for arrivals would not appeal at all. Yet, if this were done, the person arriving at the other end would pronounce this new mode of transportation a complete success and would be happy to use it for the return journey (only to find that it entails a horrible death). Meanwhile the product back at the starting point would tell all its friends what a good holiday it had and would recommend this mode of travel. But would this person be you – after all, your body ended its days being dissolved into atoms for the raw material store? And the first-generation copy met the same fate.
Yet, in real life, we have a situation not too different from that. Many, if not most, of the atoms that made up my body twenty years ago are not the same as those in my body today. But I am still the same person with the same consciousness – a little more aged and less fit perhaps, but essentially the same person. We are happy to accept that the essence of a person is the information coded by the atoms, not the atoms themselves, some of which came perhaps from some sausage eaten a couple of years ago.
So what does this tell us about the feeling of self? A brain will be aware of which body it is in and will have a strong tendency to preserve that body. That is, after all, what millions of years of selective trials have built it for. But the teleportation thought experiment shows us that the sense of self, although not illusory, is a lot less absolute that we might imagine it to be.

Consciousness
It is my contention that this discussion is sufficient to understand the question of consciousness. Consciousness, as we recognise it, is a suite of information processing functions. These components are sensory input, reduction of data redundancy, decision making to maximise an objective using internal models, memory, communication, self-preservation and, above all, the sensing and knowledge of internal mental and physical activity. There is not much mystery although many would not agree. We must ask, is there really a problem? Do those who insist that there is a deep problem actually want it to be a problem? Do they prefer to keep the mystery, so that desire can still have a licence to triumph over evidence?

Animal Consciousness
But, assuming that we know what we mean by human consciousness, how can we imagine what the term would mean for other animals? For the higher vertebrates, it probably has some similarities to our own, but with some known differences. We can conjecture this from the fact that we come from closely related evolutionary lines of descent and also from the similarities found in the structure of our brains and behaviour. Let us think what these would be in, for example, a dog.
The dog has no language, but certainly takes in information, processes it into theory models and produces actions to serve objectives. It has memories of past experiences and habits, but shows no sign of making any plan beyond today. It almost certainly experiences gratification in sex and food, but without thought of procreation or nutrition. Although very different from the human mind, it is perhaps not too difficult to imagine what it is like to “be” a dog. There are areas in which the dog’s information processing is superior to ours. The sense of smell is much better and it is easy to show that the dog must have in its brain a detailed “smell map” of the area in which it works. It is difficult for us to imagine that.
Other species are even harder to imagine. A bat has little vision, but emits high-pitched squeaks and forms a detailed picture of its environment from the returning echoes. Just try it some time. Shut your eyes and squeak and try to form a picture of your surroundings. I am sure you will be wasting your time, because although you have the vocal and auditory apparatus, you simply do not have the brain processing equipment to form the theory of the shape of environment that results (although it is said that blind people can learn to do it to some extent). But a bat can do this in enough detail to fly at high speed close to obstacles. It can even catch a moth that is taking evasive action. As it approaches a target that needs a more detailed image, its squeak frequency will increase. So the bat is our superior in one aspect of its brain, in a way that we find almost impossible to imagine, although in general terms, it the intellectual inferior of the dog.
Herbivores tend to be simpler than carnivores in their intellectual endowment. A cow has total contentment with a field of grass and the absence of danger. Its programmed objectives are not the same as ours. It is ridiculous, for example, to suppose that it might suffer from boredom, although they do sometimes exhibit curiosity. Using the same process that evolution has programmed into all organisms, it will reduce the redundancy of its incoming data stream by applying the principle that a sequence that has repeated many times is likely to repeat again. It almost certainly never realises that these benevolent bipeds that bring food and other benefits will, one day, suddenly turn into murderous carnivores.
We can be reasonably sure that many mammals have a consciousness in some ways similar to our own, but every consciousness is not the same. Consciousness is a combination of different functions and these come in different proportions. Their brains can receive inputs that we would call pain and produce the internal signal that we would call fear. They can model their environments and process these against objectives to produce decisions, but none other than the human extrapolates the model into the distant future. Even the squirrels who bury their nuts, or birds who build their nests, seem to be following a fixed routine. But they can sometimes do better than humans in some specialised tasks. Their (and our) objectives have been programmed by evolution to produce the best results in their niche and they differ greatly from us and from each other.
Mammals are perhaps easiest to empathise with because they are so similar to us, especially if they have big appealing eyes and cuddly fur, but reptiles, particularly snakes seem to have minds very different from our own. Although slugs process a great deal of information and act in a way that is better than random in the pursuit of their survival and reproduction, it is difficult for us to visualise what it is like to be a slug. Even some plants react to sensed information to produce better-than-random outputs, although they clearly lack most of the elements of consciousness.
So we must conclude that animals are conscious, just as we are, but it is a different kind of consciousness. Consciousness is made up of a suite of different functions and they are present in differing proportions and amounts in each species. There is no hard boundary between conscious beings, automata and inert objects – only a continuum measured in differences of degree.
But who says consciousness matters anyway? If we have to decide to unplug the life-support system of a person who will never recover consciousness we will give the matter much careful thought and discussion. If a fully conscious, but injured, dog is to be “put to sleep” we may agree to that with only mild regret. If a “conscious” computer of the future is to be broken up for spares, there would seem to be no moral problem at all, although we might be swayed if it produced piteous and realistic pleadings for mercy.
Yet why should we listen to a machine that produces entreaties that it be spared? No matter what signals rush around its circuits, they are only electro-mechanical functions caused by the “don’t switch off” objective that some human programmed into the machine when it was constructed. Why should I be concerned by these signals? They will certainly stop when I pull the plug and they will have no existence when I have dismantled the components and wiped the memory units. There is no convincing reason why I should not go ahead.
And then, of course, I can apply just the same argument about an animal. A human did not program the fear and pain signals rushing around its nervous system, but they were formed, for the most obvious reasons, by natural selection. Animals that produced these internal nerve impulses were more likely to survive and transmit their genes to the next generation. So it is no problem for me. All this nervous activity will cease as soon as I kill the animal and will have no existence at all by the time it is made into sausages.
The next question is now why we should not kill people using the same argument. Nothing in the above gives any reason why we should not. Another person’s consciousness, fear and pain are only electrochemical signals. They are there because they have been programmed by natural selection in exactly the same way as the other animals, so why should a different conclusion apply?
So it seems that we can go ahead and kill anything or anybody and it is no problem. If you think that this conclusion is mistaken (and I certainly do) then I think we have to abandon the idea that we base our moral duty to animals (or machines or people) on consciousness. Someone who does not share our emotional predisposition will say “so they process information to generate fear and pain signals – so what, why should I care – it doesn’t spoil the quality of the meat?” It is not easy to produce a logically compelling refutation. We will have to investigate a little further before we can find the answer.

What does philosophy offer so far?
Published philosophy is disappointing. Mary Midgley in 1983 produced a little book4 entitled Animals and Why They Matter. If you read this to obtain the answer that the title promises, you will be disappointed. It rambles on for its 145 pages, dabbling in different aspects of the subject and raising a few good points, but never producing a deduction from reasonably certain premises to produce a prescription in which we might have confidence. The final sentence at least is honest: “It is too large a problem for this book, and I must leave it to my readers to settle it.” One wonders why she would have sought to publish, or why Penguin would have consented to do so.
In A Companion to Ethics edited by Peter Singer5, a chapter entitled Animals is written by Lori Gruen. Unlike Midgley, she has definite conclusions. The chapter starts with “In order to satisfy the human taste for flesh, over five billion animals are slaughtered every year in the Unites States alone. … An estimated 200 million animals are used routinely in laboratory experiments around the world annually. A large proportion of the research causes the animals pain and discomfort while providing absolutely no benefit to human beings.” At the end of the chapter she concludes “While there are different philosophical principles that may help in deciding how to treat animals, one strand runs through all those that withstand critical scrutiny: we ought not to treat animals the way we, as a society, are treating them now.” Sadly Gruen produces, despite wandering through the ideas of a few other thinkers, no coherent argument to support her views. Peter Singer himself has written a book entitled Animal Liberation6 of predictable and similar conclusions. In it he appears to think that the fact that animals suffer means that we ought to show a moral duty to them. As Singer himself must know, an “ought” cannot be derived from an “is” and where such a deduction appears successful, there must be a value assumption that is unstated. In this case it is simply that we ought to show a moral duty to all creatures that can suffer, which is what he is purporting to prove!
So what are these arguments by which the philosophers might claim a better knowledge than the man in the street? Let us examine a few of them.

Animal Rights
Animal rights is the label that is much used today. It is an imprecise term, but Gruen quotes Tom Regan, the author of The Case for Animal Rights as saying that beings with inherent value have rights. Only “subjects-of-a-life” have these rights, and Regan believes that all mentally normal mammals of a year or more are subjects-of-a-life and thus have inherent value which allows them to have rights.

This kind of talk is full of holes. What is special about one year? Gruen seems to think that rights are things that are a natural property of an individual like its weight or colour. This is nonsense. Rights are a convention between people in which some grant (or are sometimes forced to grant) that the interests of the others will in certain circumstances prevail. The definition of who exactly holds the rights is likewise poor. As in most approaches to this problem, the question of where to draw the boundary exposes the flaw in the whole logic. If mammals have rights, how can we refuse reptiles, amphibians, insects, or bacteria? Do we have duties to plants? There are some who think we do. Consciousness as such, we have already seen to have its difficulties. But even if the assignment of rights were consistent, where is its authority? What makes it better than a statement of the author’s gut feeling?
It is a widespread approach. John Rawls has said, “The capacity for feelings of pleasure and pain and for the forms of life of which animals are capable clearly impose duties of humanity and compassion in their case.” But there is nothing clear about it; it is equally short on proof. It only makes sense when it can appeal to an unstated premise that is the same as the asserted conclusion – not a great deductive achievement!

Equality
The argument that we should regard animals as equals in our moral community is often advanced, but it is difficult to supply any reasoning to support it. The term “speciesism” has been coined to attempt to brand the opposing view with the stigma of racism, but that is name-calling, not argument.
Midgley points out that equality has never extended to an unlimited group – even some human societies professing equality have taken it for granted that slaves were not included! She sees rather, concentric circles of diminishing moral claims. That reflects somewhat our usual feelings. We would not feel as generous to a stranger as to our own friends and family; we would regret the death of a dog, but not as much as that of a person and so on. Extreme need has a widening effect. If a stranger’s life were in immediate danger, it would take precedence over the trivial needs of one’s own family. It is perhaps a more practical approach in the real world and describes better what actually happens, but it is just a short on reason and leaves large areas awaiting definition.

Utilitarianism
Utilitarianism, the greatest benefit to the greatest number, has always been an incomplete idea, even among people. In a small group, if a man chooses one unit of benefit for himself when he could instead have obtained two units of benefit to be shared around his group, he could be deemed selfish. Clearly, if all members of the group act in this way, the group will be consistently worse off on average. If they can find a way to promote unselfish choices among their members, all could gain in the long run.
We can recognise here a multi-person version of the classic co-operate/defect dilemma. As in other cases, it works best when there is reasonable hope of reciprocation. Even an approximation to a utilitarian society depends on the ability of the participants to understand the idea of a contract and reciprocation. This obviously makes the extension beyond the human species difficult.
Some, of course, commend that we should guide our actions for the greatest good of the greatest number without regard to our personal share. Experiments to make that work without enforcing mechanisms have been tried often and have never worked. We can be sure they never will, as they depend on an unrealistic view of human nature. Suppose someone is told her child has died. Suppose, instead, she is told that 1000 children have died in an earthquake on the other side of the world. Which piece of news would cause her most grief? Of course, how she will feel and how she ought to feel may be two different things.
Midgley mentions the co-operation of ants, wondering why humans cannot do the same. She seems to be ignorant of the difference in genetic mechanisms of the hymenoptera, which generates this particular pattern of altruism and is the reason why it is less likely to work in mammals. (It is amusing to note that elsewhere she reasonably questions whether ethics should extend to locusts, hookworms and spirochaetes. The ordinary mortal is left to ask, “What is a spirochaete?” It is best not to make prominent displays of biological erudition when one lacks knowledge of the basics!)

Emotion
Advocates of morality towards animals are sometimes disparagingly accused of emotion, yet we know from Humes’s law that all beliefs about value must stem from feelings. The job of careful thought, which we might call philosophy, is to try to understand our own feelings and make the decisions that emerge from them wisely.
Certainly we can be misled by unconsidered emotions. We like cheap eggs but we object to battery poultry farming. Extremists demonstrate against animals being used for research to find cures for serious illnesses, but do not demonstrate when they are made into pies. Our emotions are notoriously inconsistent. We are affected by social convention – attitudes to certain treatments of animals have changed within a few decades.
Many would feel emotional revulsion while watching a surgical operation, yet it may well be a very good thing that it be carried out. It might be a wise precaution, however, for those of a tender disposition to remain outside operating theatres. Some would say that the same thing should be said of slaughterhouses, but we must study a little further before we can hope to decide that question rationally.

Duties to non-sentient things
In a recent article in The Philosophers’ Magazine, if I remember it correctly, the author managed to arrive at the conclusion that the human race should desist from exploration of other planets to avoid damage to them. He managed to arrive at the conclusion that the pristine current state of these planets had a value that was quite independent of the interests of any persons. I know of no argument that could show that this makes sense, and the author certainly did not present one. Similar arguments occur for the conservation of nature and of old buildings or other human artefacts. Biodiversity is promoted as a goal in itself, yet it is difficult to say why.
Most of these goals can be established without recourse to disembodied values that “exist” without anyone holding the value. Certainly the destruction of things that cannot be recreated can be opposed rationally, on the obvious grounds that a choice may be lost which later discoveries may reveal to be valuable to humans.

The Contributions of Philosophy
The contributions of philosophy, we may say, are disappointing. Most of what has been offered fails to produce a compelling argument and leaves our attitude to animals little altered from our untutored prejudices or the fashionable view of our time. We feel more protective toward animals that are appealing in appearance, although why the cuddly ones are appealing is an interesting psychological question. Those with a bad public image, such as rats, receive scant sympathy despite a probably equal capacity for suffering. We have no idea why we value humans more than animals – we just do. In this case, as in so many others, the fishwives are just as good as the philosophers.

An Evolutionary Approach
A long time ago there was no sign of life on our planet – only inert molecules that were undergoing change under the laws of physics and chemistry. At some point, some of these molecules started to act as templates on which others formed, and so replication began; after vast stretches of time, under the influence of natural selection, they increased in complexity. Much later, they produced animals that take in information about their surroundings and act conditionally upon it, so as to increase their selective advantage. These capacities for conditional acts were the first examples of values, choices or preferences that appeared on our planet. We ourselves are among the end products of this evolutionary process and it is from that, and from interactions between each other, that our brains and their inherent values have come.
Nowhere is there any “absolute” (if I may call it that) basis for any kind of values, yet it is easy to see why values will exist in any animal that has even the smallest amount of information-processing capacity. Those who choose outcomes that lead to above-average survival and reproduction will be positively selected and will become more numerous in the population.
If this account of the origin of our mental apparatus is true (and there can be no realistic doubt that it is true) then it must have profound consequences for any philosophy of value and ethics (and for any theory of knowledge also). I remain permanently puzzled at the failure of most philosophers to consider it seriously. By failing to do so, they reject the essential clue to solve the mystery.
For example, John Harris, Professor of Applied Philosophy in Manchester, has produced some very competent and useful work in the area of medical ethics. Yet in his book3, The Value of Life (p.186 of 255) he says, “Indeed, if one or other of the versions of natural selection describe a natural evolutionary progress for human beings, then again, while it may be evolutionarily successful, one could hardly describe the survival of the fittest and its corollary, the destruction of the weakest, as a humane (albeit a human) arrangement”. This is the only reference to natural selection in the entire book.
His reasons for considering that evolution is irrelevant are similar to those of most of his colleagues. Having understood the basic mechanism, he swallows whole the oversimplified red-in-tooth-and-claw image. Because that does not reflect his own ethical sentiments, he departs without looking further. Yet where does he suppose his sentiments have come from? We have descended from a long line of social primates whose very survival has depended on the maintenance of co-operation. Natural selection has programmed into our brains a very detailed structure of ethical instincts, a sense of natural justice and other adaptations to co-operative living. This complex suite of instincts is too comprehensive and consistent in its effects to have occurred by chance. Yet we have not obtained it from writings in the sky somewhere, nor from tablets of stone on a mountaintop. Natural selection, acting in an arena of interaction between people, is the only source that could have assembled this complex body of information.
Other animals, for example a carnivore of solitary habit, will have been programmed by natural selection quite differently. I do not think a tiger would ever think of becoming a vegan for ethical reasons, yet it might look after its own young with great tenderness.

The Central Problem of Values and Ethics
There is a fundamental problem of values. We can understand that our desires and our ethical sentiments have come from our evolution as social animals. We can observe that they are, for the most part, closely optimised to serve our survival and reproduction. We can observe that some of them no longer do so in modern conditions, although they did at one time. Should we discard those that are misfiring in this way? We can observe that natural selection has been the only source of information that has assembled us, including our sentiments, from random molecules. Yet this does not quite answer the question – how ought I to feel and to behave now? You cannot derive an “ought” from an “is”. How our instincts are, is not necessarily how they ought to be. And yet there is no other source of information on which we could rely to modify them.
This is an application of Hume’s Law, the idea that we can never say how we would like the world to be, by reasoning only from knowledge of how it, in fact, is. We are stopped short of a complete solution and yet we can show that no more complete solution is possible. It seems unsatisfactory, yet we can take satisfaction that our understanding is worlds ahead of most students of ethics who simply interrogate their own instincts and try to resolve the contradictions that appear.

The Sources of Altruism
Despite the difficulties posed by this Central Problem, it is still useful to understand why we feel instincts of altruism at all toward animals, or even people for that matter. On the face of it, another human being is a competitor for resources and is made of exactly the right mix of nutrients required to form human bodies. If Harris is right in his assessment of evolution we should have the instinct to be murderous cannibals every day. Yet we do not find that this is so, even in the worst of our fellows. It takes only a moment’s reflection to understand why someone with such instincts in the prehistoric human tribal context would not leave many descendants! Let us take a few moments to investigate what sort of ethics might be produced by natural selection.
If we are to throw light on our duty to animals, we must begin by trying to understand why we suppose we have a duty to humans. Pure unrewarded altruism, the sort that might give away all one’s money to strangers in the street, could not be stable under natural selection and for that reason is seldom seen. A great deal of human altruism is seen, however, and there are three types that are stable under natural selection. These are kin altruism, reciprocal altruism and display altruism.
It is easy to see why kin altruism is selected. Taking care of one’s own children and other relatives will obviously tend to make one’s genes more frequent in the gene pool. But altruism to strangers is also stable, when there is hope of reciprocation. It has been extensively explored in the two-person case in the game-theory analysis of the co-operate-or-defect dilemma (prisoner’s dilemma).
The real world is, of course, much more complex than the simple two-person dilemma, but the principle carries over. It often happens that an individual who grabs the maximum benefit for himself would cause a loss of benefit to others that is greater than the amount he has gained. The total benefit to the group, and hence the average benefit of its members would thus be reduced. Clearly, any group that can persuade its members to act unselfishly will enjoy an advantage over those who do not.
But biologists today have very little confidence in group selection, because a few selfish individuals could prosper, enjoying the contributions of their fellows, even while the group declines. The selfish would become more numerous and would thus be the stable form under natural selection.
Yet the prize of the greater benefit of co-operation is there for any species that can find a way to attain it, and to a large extent the human species has attained it. It has done so by devising mechanisms to control the cheats. Much of the structure of modern life consists of these. Laws, money, etiquette, fences, padlocks, policemen, taxation, revenge, gratitude etc. are all serving to prevent defection from co-operation. Co-operation becomes more difficult as group size increases and it is only the increasing development of these mechanisms that has made the modern nation state possible.
Co-operation, to gain access to the greater benefits that are possible, is clearly a form of altruism that natural selection will reinforce, provided the cheats can be controlled. But kin altruism and co-operation do not account for all of the altruism that we see. The third mechanism is display altruism or reputation building.
It is a matter of common experience that behaviour is changed when others are watching and this is not confined to picking one’s nose or making a generous donation at a charity ball. A large part of our cheat-detection mechanism is to observe the character of our fellows, not only in their dealings with us, but also in their behaviour towards others. This information is very useful when trying to judge whether a person will be likely to reciprocate favours, keep promises, or otherwise be worthy of trust. It is clear that evolution will create an instinct to observe people; equally, it will create an instinct to behave well when being observed.
Display altruism is a very important part of human behaviour and is a major component of the unselfishness that is stable under natural selection. It is part of the means by which the two-person dilemma model can be generalised to a multi-person model, but it is important and distinct enough to be given a separate name. Many complications are worthy of further exploration – for example, display altruism seems stronger when the observer is of the opposite sex and might have even a small probability of being a potential mate. Display altruism can be laughably observed in many writers, teachers and students of philosophy, who often imply the highest moral principles in themselves while discussing ethics. By doing so, they cloud their view of the mechanisms and make a clear-headed analysis more difficult.
These forms of altruism, which we can show to be evolutionarily stable, do indeed correspond closely to the way humans actually behave and it is easily possible to think of other “saintly” forms of altruism that are unstable and, as a result, are never or seldom seen. Let us not deceive ourselves; the altruism that we “just feel” has been installed in us by the natural selection of our ancestors and there is no other source of this detailed information.
Of course there is a gap between recognising where our moral sentiments have come from and saying that is how they should be. Yet, if we are to debate this, we must not forget that the feelings we are bringing to bear on the question have come only from this one source, and, following Hume, only our feelings can ultimately guide us. It is beyond the scope of this paper to fully examine this question. I have tried (no doubt inadequately) to explore it elsewhere1.
Now there is something else that we should observe about the “is” as opposed to the “ought” of our moral sentiments. The modern world has brought advances in information technology. The most important step was the invention of printing by Johannes Gutenberg in 1450, but we can observe how it is still accelerating in our own time. Radio, television, the Internet, mobile phones etc. have all increased information flow. Information is now so easily transmitted that there is little limitation in our ability to obtain it; the limitation is in our ability to absorb it.
The effect of this explosion of information is to alter our traditional environment for display altruism. In the early 19th century, the landlords in the Highlands of Scotland realised that sheep farming was much more profitable than collecting what rents could be squeezed out of crofting families. So they gave them notice, then appeared with bailiffs, destroyed their crops, set their houses on fire and forbade them to re-enter the land. Extreme destitution was caused and the descendants of these people (or at least those that survived) are now spread in North America, Australia and elsewhere. Now let us ask, would the story have been the same, if a News-at-Ten camera team had been present to record the proceedings? Examples are many. Slavery, burnings-at-the-stake of unbelievers and many other unpleasantnesses have been suppressed by communication technology.
But the system is not perfect. A recent story is of a 60-year-old hostage taken in Iraq and tantalisingly shown on videos before being murdered. The media coverage was great, and the highest in the land did all they could and repeated their distress. Large public displays of mourning took place. In the same week two young soldiers died, presumably a greater loss, but were almost unnoticed. To the media, your life is much more important, if you are “interesting”.
So, after this tour around the sources of our altruistic instincts, let us return to the problem that we are trying to solve – what about animals?
The first kind of altruism, kin altruism, cannot be applicable here, but the other two, reciprocal co-operation and display altruism, may well play a role. Co-operation is certainly a possibility. Good treatment of an animal will create in it a confidence that will lead it to behave in a way that its human trainer wants. Even the simple conditioning of “do the trick – get the reward” is a form of reciprocal co-operation.
But by far the greatest part of our instincts that lead us to treat animals well is display altruism. Kant observed "He who is cruel to animals becomes hard also in his dealings with men. We can judge the heart of a man by his treatment of animals". This sums up the common attitude exactly.
But Kant (quoted by Midgley) also says, “So far as animals are concerned we have no direct duties. Animals are not self-conscious and are there merely as a means to an end. That end is man. We can ask, ‘Why do animals exist?’ But to ask ‘Why does man exist?’ is a meaningless question.” In this he is wrong, wrong, wrong. He is wrong that we have no duties; wrong that animals have no consciousness, wrong that their end is man and wrong to say that discussion of the reason for human existence or purpose is meaningless. To a man the dog may only be a means to human ends, but to the dog, it is the man who is a means to his ends. This whole assertion illustrates how weak philosophy was then, and to a large extent still is.
But his point about display altruism is correct (although he would not have called it that). Suppose that a young woman observes a young man who takes a delight in torturing his dog. It would be a reasonable inference that he might not make an ideal husband. We can see why an instinct to disapprove of cruelty would evolve in young women and why an instinct to desist from it (at least while being watched) would evolve in young men. Of course this is not the only situation in which the conclusion should apply. It would be similar if the sexes were reversed or even if mate selection had nothing to do with it. For any kind of co-operation, cruelty would be a good predictor of an untrustworthy partner.
The same process will apply to cruelty to people. A visibly cruel person is likely to be a bad co-operation partner, but with humans there is a more important element. The co-operation that is special is the one in which we agree that we should not kill or harm each other. This is a very beneficial co-operation because the benefit we enjoy from not being killed is much more valuable than the satisfaction we might get from killing people who annoy us. It is a good trade, and we would be much worse off if we did not make it. Of course there are deplorable exceptions, but the fact that we deplore them is witness to the fact that we do not simply enter this co-operation by negotiation. Because it has always been the best strategy to enter it, we are programmed by natural selection to do so.
An interesting situation is raised by the discovery of Homo floresiensis, a different species of human found to have been living on the island of Flores only a few thousand years ago. This creature was only about a metre tall, but was a toolmaker and hunter. It was definitely human and not ape and had diverged from the human evolutionary path comparatively recently. There has even been speculation that some living communities of them might survive today in remote locations. The ethical question would then be whether we would view such a close relative as human, qualifying for the special ethics that we reserve for humans, or as animals worth conserving as an endangered species. I am sure we would feel it was wrong to farm them for their meat. There is no precise answer, except to note that this is not a hard boundary – it gets fuzzy at the edges.
Desmond Morris reflected on this discovery and the idea that humans are different in nature from other animals. “Darwin put a stop to this nonsense with his theory of evolution, but amazingly the blindingly obvious truth he discovered is still resisted by large sections of the human population. They stubbornly continue to insist that we are some kind of special creation. The arrival of ‘Mini-Man’ is going to give them nightmares. How can he be ‘semi-special’? That won't make sense. He can't very well have a semi-soul. So Mini-Man might just be the evolutionary jewel that, once and for all, sets human beings firmly in the animal kingdom, where scientifically they belong.”
But despite the fuzzy boundary, we can see why there is a distinction between the duty that we accept to people and that to animals. Both are made selectively stable by display altruism, but towards humans we have a direct and vital reciprocal element. We can thus understand why we give a great deal of thought before switching off the life support machine on a brain-dead patient, who has no hope of recovery, whereas a conscious race horse with a broken leg may be shot because it is not economic to care for it. We can also understand why some people get worked up irrationally about the abuse of stem cells (which are not cuddly with big eyes) and are less concerned with cats and dogs (which sometimes are).
This distinction between our inherited sentiments toward people and animals is similar to Midgley’s concentric circles, but rather than simply propose these as a description of subjective feelings, we can understand where they have come from. So we are left with display altruism as the main reason why we must show animals some kindness. Some will say this is too cynical because they “just feel” a revulsion against cruelty. They are right to feel that and the actions that emanate from their feelings will be the right ones. But it is the job of the philosopher to understand a little more deeply what is going on. When we examine what happens backstage, it may not look as pretty as the view from the dress circle.

Conclusions:
So can we reach a conclusion that will be any more certain than Midgley’s invitation to think for oneself? I think we can certainly reach a deeper level of understanding than she offers, yet we are still hesitating before the Central Problem. To understand why we feel as we do is not the same as proving that we ought to feel that way. Yet there is no other evidence on which we can draw to say how we ought to behave. Perhaps I have to imitate Midgley here and say decide for yourself.
But, if we can say that we will decide based on how we feel, but using the evolutionary understanding to help us toward understanding and consistency, I believe we will find ourselves in a better position to make prescriptions. With this in mind, I would offer the following practical points, which are no more than my own decisions:

1. It is correct that we should value human life highly and more than that of animals.

2. The boundaries of human life are, to some extent, arbitrary. The important thing is to maintain the sanctity of our co-operation not to harm each other. Thus abuse to stem cells is only a problem if we choose to make it one. We should choose our arbitrary definitions to bring the greatest benefit to humanity.

3. It is acceptable to use animals for food, clothing, other products and for experiments.

4. In animal husbandry, great care should be taken for the animal’s “welfare”. Its exposure to fear, pain and suffering should be prevented.

5. Our duties to wild creatures should avoid significant human-inflicted suffering, but it is no part of our duty to supervise wild carnivores to train them to become vegetarians.

6. We should make efforts to safeguard the environment, the natural habitat, the biodiversity of the ecosystem and endangered species. But the only reason for this that we can convincingly demonstrate is the benefit of our fellow humans; nothing more.

References

1. Cameron, Donald: The Purpose of Life (2001) Woodhill

2. Cameron, Donald: Information Theory and Entropy – Their relevance to Philosophy (2004) Proceedings of the Bath Royal Literary and Scientific Institution Sept 2003 – August 2004.

3. Harris, John: The Value of Life (1985) Routledge

4. Midgley, Mary: Animals and WhyThey Matter (1983) Penguin Books.

5. Singer, Peter, ed.: A Companion to Ethics (1991) Blackwell.

6. Singer, Peter: Animal Liberation (1975) Random House