Animal Experiments

The medical-benefit defence

The classic modern case was made by the philosopher Carl Cohen in the New England Journal of Medicine (1986). Animals have no rights, he argued, because rights arise only among beings who can make moral claims on one another; we do have obligations not to cause them needless pain, but research that relieves human disease is not needless. And the balance of consequences favours it: insulin, polio vaccines, antibiotics such as penicillin and many surgical techniques were developed or tested with animals, and most regulators still require new drugs to be tested on animals before they are given to humans. Cohen also defended what Singer calls speciesism: giving greater weight to human interests is not a prejudice but “essential for right conduct”. Earlier defenders drew a narrower conclusion. Bentham saw no objection to painful experiments on animals “when that experiment has a determinate object, beneficial to mankind, accompanied with a fair prospect of the accomplishment of it” (1825), and Darwin, who supported legal limits on vivisection, wrote in 1881 that whoever retards the progress of physiology “commits a crime against mankind”.

Held by

• Jeremy Bentham1748–1832
• Charles Darwin1809–1882

Replace, reduce, refine

Kant already held that agonizing physical experiments “for the sake of mere speculation, when the end could also be achieved without these, are to be abhorred” (Metaphysics of Morals, 1797, § 17). The zoologist William Russell and the microbiologist Rex Burch turned this kind of thought into a programme (The Principles of Humane Experimental Technique, 1959): replace animals with other methods wherever possible, reduce their number to the minimum needed for a sound result, and refine procedures so that they cause as little pain and distress as possible. The “3Rs” now underlie laws and guidelines in many countries, among them EU Directive 2010/63/EU, which also requires every project to be authorised and every procedure to be classified by severity. In the same spirit the EU banned testing cosmetics on animals in stages, completed by a marketing ban in 2013, and in the United States the FDA Modernization Act 2.0 (2022) ended the legal requirement to test new drugs on animals. The position is pragmatic: it accepts that some experiments are still needed, but treats each one as a cost that must be justified and, as soon as possible, avoided. It draws criticism from both sides – from researchers, who fear that harm–benefit tests hold back basic research whose value cannot be known in advance, and from animal advocates, who object that the 3Rs regulate the use of animals without questioning it.

Held by

• Immanuel Kant1724–1804

The argument from species overlap

Singer puts a question of consistency to experimenters (Animal Liberation, 1975): would they be prepared to perform their experiment on an orphaned human infant with severe and irreversible brain damage, if that were the only way to save many lives? If not, while they are prepared to use a monkey or a dog that is at least as aware and as capable of suffering, their choice rests on species alone. Because some humans have capacities no greater than those of some animals, any criterion that excludes all animals from equal consideration either excludes some humans as well or comes down to membership of the species. The argument was long called the argument from “marginal cases”; following Oscar Horta (2014), it is now often called the argument from species overlap. Singer’s conclusion is not an absolute ban: an experiment that would really save many lives might be justified – but then, by the same reasoning, so might one on a human being with comparable capacities. Experiments that no one would think justified on such a human, which on his view includes most of those now performed, are not justified on animals either.

Held by

• Peter Singerb. 1946

No harmful use, whatever the benefit

For Regan the size of the benefit does not settle the question (The Case for Animal Rights, 1983). Animals used in research are subjects-of-a-life with a right not to be harmed for the sake of others; to use them as means to human health treats their value as if it lay in their usefulness, which we would not accept for any human research subject. The rights view therefore calls not for better conditions in laboratories but for the total abolition of harmful research on animals – “not larger cages, empty cages” – even where it might bring great benefits. Its defenders accept the cost: some knowledge will come later, or not at all. They compare this with the ban on experiments on humans without their consent, which also slows research and which almost no one now regards as too high a price.

Held by

• Tom Regan1938–2017

Poor predictors? Translation and alternatives

A further argument is empirical: animal experiments are not only ethically costly but often poor guides to what happens in humans. About nine in ten drugs that enter clinical trials fail there, most often because they turn out not to work or to be too toxic, although they had passed animal tests (Sun and colleagues, 2022); an umbrella review by Benjamin Ineichen and colleagues (2024) found that about half of the therapies tested in animals moved on to human studies, but only 5 per cent were approved. Defenders of animal research point out that the same review found 86 per cent agreement between positive results in animal and human studies, and that failures also have other causes, among them poorly designed studies on both sides. Meanwhile methods without animals are developing quickly – cultures of human cells, organoids and “organs on chips”, computer models and better use of human data – and in 2025 the US Food and Drug Administration announced a plan to phase out animal testing requirements, starting with monoclonal antibodies. Supporters see a chance for research that is both more humane and more relevant to humans; sceptics stress that effects on a whole organism, such as on the immune system or in pregnancy, cannot yet be studied reliably without animals.