Read more: Special report: Morality put to the test

Some philosophers say intuitive moral responses are what count – but evidence on the nature of morality undermines this authority, says Peter Singer

The Enlightenment philosopher David Hume pointed out long ago that no combination of statements about what “is” the case could ever allow one to deduce what “ought” to be. After all, in deductive arguments, the truth of the conclusion is already contained in the premises. But when scientists deal in morality, deducing an “ought” from an “is” is often precisely what they attempt to do.

Fortunately, a new generation of scientists has emerged who seek to shed light on morality without attempting to annex the entire field. They approach ethics from several directions. For example, primatologist Frans de Waal has observed the rudiments of ethical behaviour in non-human primates, which appear to understand reciprocity and may have an elementary sense of fairness. Psychologist Paul Bloom has found prototypes of our moral judgements in babies as young as 3 months (see page 42). These examples provide powerful evidence for an innate component to our moral judgements.

Social psychologist Jonathan Haidt found that people are often unable to give reasons for their snap moral judgements – he calls this “moral dumbfounding” – suggesting that many are based on intuition. The reasons we offer tend to be rationalisations. As Haidt puts it, the emotional dog wags the rational tail.

Emotion vs Cognition

Moral psychologist Joshua Greene’s work supports this conclusion. He used fMRI to see what happens in our brains when we respond to moral dilemmas. His subjects were told they could push a person in front of a train in order to save five others, or throw a switch to the same end. Almost everyone believed throwing the switch permissible, but most thought it wrong to push someone. Greene found that those who object to pushing showed greater activity in parts of the brain associated with emotion than they did in parts of the brain associated with cognitive processing. By contrast, those who were willing to push showed more cognitive activity.

Greene interprets the results as indicating that we have evolved automatic negative responses to hands-on violence, but not to throwing switches. When our decisions are freed from automatic responses, we are more likely to base them on the consequences of our actions.

So let’s assume that we have instinctive moral responses to a variety of situations of the kind encountered by our ancestors throughout our history, though modified by our culture and upbringing. What would follow from this about what we ought to do?

It certainly doesn’t follow that we ought to do what our instincts prompt us to do. That might have enhanced our survival and reproductive fitness in an earlier period, but may not do so now; even if it did, it could still be the wrong thing to do. Consider the ethics, for example, of having a large family on an overpopulated planet. Rather, by undermining the authority that some philosophers have given to our intuitive moral responses, the new scientific lines of evidence about the nature of morality open the way for us to think more deeply, and more freely, about what we ought to do.

“What once enhanced our reproductive fitness might now be the wrong thing to do”

Consider, for example, the idea that all human beings have a moral status superior to that of any nonhuman animal. Such a belief flatters our pride, and provides a convenient justification for eating animals as well. It seems to be supported by scripture. Yet once reason – part of which is, of course, our understanding of evolution – leads us to abandon belief in the literal truth of Genesis, this moral belief becomes impossible to justify. Unless humans have a divinely bestowed right of dominion over other animals, how can an anencephalic baby, doomed by the absence of a cortex to a mere vegetative existence, have a moral status superior to that of a chimpanzee?

Rejecting the idea of a sharp moral divide between humans and animals, however, subverts not only our assumed right to exploit animals, but also the traditional ethic of the sanctity of human life, which underlies the heated issues of abortion and euthanasia. If you want to blame the use of reason for this, however, you need to suggest a better way of deciding on moral issues. We cannot just stand by our conventional morality, as if that provided some guarantee of not going astray. If we had done that a century ago we would still be discriminating against racial minorities and women.

Nor can we rely on our feelings, for they are an unreliable guide. Hermann Göring famously said: “I think with my blood.” It isn’t good enough to respond: “Me too, but my blood thinks differently from yours.” Reliance on what one intuitively “feels” to be right has led many people to support not only racism and sexism, but also burning witches and persecuting homosexuals.

Evolved morality

Recent research suggests that there may be, on some matters, a universal moral sense: virtually every human society recognizes obligations to kin, has a taboo on at least some forms of incest, and has a sense of reciprocity, or fairness. A few incautious scientists have used this research to suggest that somehow we can derive the true or best moral code from our knowledge of evolutionary psychology.

Such attempts inevitably founder on David Hume’s well-known distinction between the “is” of science and the “ought” of normative ethics. Knowing that a widespread moral “intuition” is the result of evolution does nothing to support that intuition, it merely shows that we evolved in conditions in which having this intuition was advantageous – if not to individuals, then to the genes that gave rise to them. That doesn’t tell us whether these intuitions have good effects today. If anything, understanding that what we take to be a moral intuition is the outcome of natural selection undermines the sense we might otherwise have had, that our moral intuitions provide insight into an objective realm of moral truth.

Similarly, the fact that we, or many of us, lack an intuitive feeling that an action is right is perfectly compatible with the action being right, or even obligatory. Many people deny, for example, that we have any obligation to help distant strangers. Our intuitions about helping people who are not conspicuously present to us – for example, those living in great poverty in Africa – are much weaker than our intuitions about helping someone we know, or can see in front of us. That may be linked to the fact that helping distant strangers in need is not, in evolutionary terms, a good strategy – it won’t help you to survive and reproduce. But why should we take that as a ground for denying that we have an obligation to help distant strangers? Don’t they suffer just as much as those close to us?

Reason or nothing

So there is no alternative to using reason in ethics. That’s especially true in an era in which the progress of science constantly brings up new ethical questions: about the use of embryos to obtain stem cells, about genetic testing and selection of our offspring, about whether the developed nations ought to take the lead in reducing their emissions of greenhouse gases. How, without using our reason, could we find answers to such ethical questions?

That doesn’t mean that reason alone is enough. Kant thought it possible to derive moral principles from pure principles of reason, but not many philosophers agree with him now. Whatever the right answer to that long-running debate may be, reason can and should play a critical role in our moral decision-making. When we try to explain what grounds our particular moral judgements, we find that many of our intuitions are at odds with each other, and that others are based on arbitrary and indefensible distinctions. Perhaps reason doesn’t tell us that there is just one true moral view, but it does tell us that there are several moral views that are wrong. That’s an important beginning. We need more reason in ethics, not less.

My correspondent maintained that scientists really cared for the animals
they used, and defended only essential research. I replied by asking the
researcher to consider the following experiment, the details of which were
published in a paper in the journal Radiation Research. At the US Armed
Forces Radiobiology Research Institute, in Bethesda, Maryland, researcher
Carol Franz spent nine weeks forcing 39 monkeys to run on a cylindrical
treadmill called an ‘activity wheel’. If they did not run for the prescribed
periods, they got an electric shock.

The monkeys were subsequently irradiated; those receiving higher doses
vomited repeatedly. All the monkeys were then put back into the activity
wheel. The shock intensity was increased to 10 milliamps – an extremely
intense electric shock. Some monkeys continued to vomit while in the activity
wheel; they took between one-and-a-half and five days to die.

I invited the researcher to show his concern for animal welfare by urging
the United States Government to stop such experiments. I have had no reply
to my suggestion.

In slightly more civilised countries, such as Britain and Australia,
we have laws and regulations that should suffice to prevent the occurrence
of such experiments. In Australia a new Code of Practice for the care and
use of animals for scientific purposes has just been released. Accepted
by all Australian universities and major research institutions and funding
bodies, it requires animal experimentation ethics committees to review all
proposals for animal experimentation, and takes as its guiding philosophy
the idea that, unless there is evidence to the contrary, the committee must
assume that animals experience pain in a manner similar to humans.

Enlightened laws are better than unenlightened ones. But even the best
codes do not deal with the underlying problem: the institutional practice
of animal experimentation makes it difficult to think of animals as anything
but laboratory tools.

I have just been shown over a laboratory near Melbourne. It was as well
run as any animal research facility I have seen. One room we saw contained
two rabbits. They were living biological factories, producing antibodies
used in other research.

The labels on the cages indicated that one of the rabbits had been there
since November 1988, and the other a year less. So they sat, as the months
stretched into years, each lying on a wire floor in an individual square
cage with sides perhaps twice as long as their hunched up bodies. The cages
were on shelves, together with another cage containing two guinea pigs.
The rest of the room was empty.

The scene suggested that in all that time, no one had really thought
about the animals as living beings with a need to do more than merely sit
on a wire floor and produce antibodies. Anyone who had done so might have
realised that the rabbits would have been more comfortable living in a pen
on the vacant floor, with hutches and some straw. More troublesome to clean?
A little, maybe, but it all depends how much we care for the rabbits’ well-being.

That is one way in which good regulations are inadequate. A different
defect in the regulatory approach has recently received wide publicity in
Britain. The National Institute for Medical Research at Mill Hill, in North
London, is the largest laboratory directly funded by the Medical Research
Council, and might be assumed to have high standards of animal care. Yet
a film of research carried out there by Wilhelm Feldberg shows rabbits being
cut open under anaesthesia that is, as an assistant put it, ‘a little on
the light side, possibly’.

In the film, the animals are visibly squirming, and one rabbit undergoing
surgery is seen struggling vigorously to get to its feet. The Medical Research
Council is investigating whether these experiments breached either the law,
or standards of ‘good practice’ in animal experimentation. But without an
observer at the elbow of every experimenter, it is difficult to see how
such occurrences can be eliminated. The balance of power in a laboratory
is too unequal.

We know what happened when one race of human beings used members of
another race as living tools; but we still seem to think this attitude appropriate
when it is applied to members of another species. Animals can be ordered
from commercial suppliers as routinely as we may order glassware. The instruments
by which they are made to suffer are promoted like any other gadget.

Consider this handy piece of research equipment which Nature introduced
in 1988 to its readers in its ‘New on the Market’ section:

The latest animal research tool from Columbus Instruments is an air-tight
animal treadmill for the collection of oxygen consumption data during exercise.
The treadmill has isolated running lanes with separate electrical shock
stimuli which can be configured for up to four rats or mice. . . The basic
Pounds sterling 9737 system includes a belt speed controller and an adjustable
voltage shocker. The Pounds sterling 13 487 fully automatic system can be
programmed to run consecutive experiments with rest periods in between,
and automatically monitors the number of trips to the shocker grid, time
spent running, and time spent on the shocker grid. (Nature, Vol 334, page
454).

Note that this is not a one-off contrivance, but equipment in sufficient
demand to be manufactured in different models. The existence of such a machine,
and the manner in which it is presented to the readers of one of the world’s
leading scientific publications, is eloquent testimony to the failure of
the scientific community to grapple seriously with the moral issue of the
status of animals.

Because animals are not things for us to use, the animal liberation
movement will remain dissatisfied with conventional ‘high standards’ for
laboratory animal care. Instead it will seek to extend the basic moral principle
of equal consideration of interest – which we now apply to all human beings
– to nonhuman animals as well.

Peter Singer is director of the Centre for Human Bioethics at Monash
University, Australia. He is the author of the influential Animal Liberation,
first published in 1975. A second, extensively revised edition, has recently
appeared in the United States, and will shortly be published in Britain
and Commonwealth countries