Professor Frey expresses surprise at my assertion that, ultimately, nothing is provable in ethics. What about Pol Pot’s atrocities, he asks—surely we can all condemn them? Let us take a more recent example, the Beslan school massacre. The terrorists appear to have weighed against the unquestionably serious harm to the children (and adults) the boost to their cause which they judged the attendant publicity would achieve. I may believe that even to attempt such a utilitarian assessment is obscene, and most would no doubt agree. I cannot, however, prove that this is the wrong approach. A few people, at least, take a different view.

The example shows the problem with utilitarian assessments where the harm is both certain and direct, the victims and beneficiaries are wholly distinct and the victims have no say in the assessment—precisely the pattern with vivisection.

I do not, of course, advocate an ethical free for all. As I suggested in my article, an ethical response must be informed, intellectually coherent, and above all, consistent across similar issues. Empathy and consent are key factors: hence my comparison of animal experiments with non-consensual experiments on people.

Professor Frey’s foray into the apparently unconnected issue of euthanasia is instructive. The key to voluntary euthanasia, which I personally support if rigorous safeguards are in place, is the informed consent of the patient. The practice is a form of suicide, an act of free will, albeit one requiring the assistance of another person. Consent is crucial with permissible experiments on people, too, and the absence of even presumed consent is, I have suggested, what makes animal experiments ethically illegitimate.

Involuntary euthanasia is much more problematical. But if it is ever legitimate—where a baby’s suffering is enormous and will be never ending, some argue—this is surely because it is what we would wish for ourselves in that situation. In other words, we would empathise with the baby and presume his consent to the ending of an intolerable life.

Professor Frey says he is not a speciesist and indeed shows it by suggesting that the life of an anencephalic child is less valuable than that of a healthy dog. The ethical key, he goes on to say, is not sentiency but “the capacities and abilities that go towards enriching a life and so toward giving it a quality and value”. But why? Take, on the one hand, someone who helps thousands through his work and who leads a culturally rich life and, on the other, a drunken slob who beats his wife and children and spends his days watching soaps. We might well say that the life of Renaissance Man has more quality and value than that of Uncouth Brute, but would it follow that the former could with ethical approval cause harm to the latter for his own benefit? Surely not. The original Renaissance Man, Leonardo da Vinci, was opposed to vivisection, incidentally.

Professor Frey is correct that I do not base my opposition to animal experiments on a concept of animal rights. This is because ethics is about human behaviour, and ultimately rights are only accorded, whether to people or animals, if it is considered that freedom of action should be curtailed so that the group in question is protected from harm. Rights are merely the flip side of someone else’s obligations.

I maintain that the best guide to the obligations we should assume to a potential victim is to put ourselves in his or her shoes, particularly where the harm is direct and serious. Empathy is a multifaceted concept and has generated a significant body of research among behavioural psychologists, but as Martin Hoffman and others have argued, it is closely linked with moral precepts. This is a philosophy based on intellectual rigour, not emotion as Professor Frey appears to think, although it does focus on the feelings of the putative victim.

Ultimately, I cannot prove it is wrong to cause pain (or other serious harm), without any form of consent, to an animal for the perpetrator’s own benefit, any more than I can prove that it is wrong to cause pain to another person for the perpetrator’s benefit. Even religious based ethical beliefs are not empirically provable, at least in this life. I am entitled, however, to expect a consistent response to comparable situations.

There are, as Professor Frey indicates, exceptions to every moral principle—in the context of causing physical harm to others, self defence, principally—but they have to be applied consistently, not opportunistically. Ethics should not depend on what the group with power can get away with. That is precisely what vivisection involves.