LNAT model essay: animal research

A balanced answer that still takes a side, and makes its conditions do the work instead of hiding behind them. 535 words.

1. Using animals in medical research can be right, but only when there is no other way to get the knowledge and the suffering is kept as low as the research allows. Those conditions are not a hedge: most of the moral argument turns on whether they are met.

   NoteThe writer anticipates the charge of sitting on the fence and answers it in the second sentence.

2. The case for research is the scale of what it has achieved. Vaccines, antibiotics, insulin treatment and safer surgery all relied at some stage on work with animals, and so does much of veterinary medicine. Most of the animals used are mice and fish bred for the purpose; the dogs and monkeys that dominate public debate are a small share. A new drug that has never been tested in a living body is a risk to the first people who take it. If animal studies make those first human trials safer, the gain is measured in lives, not convenience.

   NoteConcrete examples of benefit, with the point that animals also gain. The last sentence states why the benefit counts morally.

3. The alternatives are improving, but they cannot yet do the whole job. Cell cultures and computer models can show how a drug acts on one kind of tissue, but not how it behaves in a whole living body, where organs, blood and the immune system interact. That is why most new medicines are still tested in animals before they are given to people. Animal studies are imperfect guides, since many drugs that work in animals fail in people; but that is a reason to improve them, not evidence that we can do without them. Until the alternatives can answer those questions, stopping animal research would mean either giving up new treatments or testing them first on human beings.

   NoteA second reason that explains why the benefit cannot simply be had another way. It sets up the condition in the opening: no alternative exists yet.

4. Against this, animals can feel pain and fear, and they cannot consent. We would never test a drug on an unwilling person, however many lives it might save, so it seems inconsistent to do so to a dog or a monkey. If the ability to suffer is what matters morally, the species of the sufferer should not change the answer. Some philosophers argue exactly this, and the argument cannot be dismissed simply because its conclusion is uncomfortable.

   NoteThe objection is put in its strongest form, as an argument from consistency, not as an appeal to feeling.

5. This objection shows that animal suffering counts, but it does not show that it counts equally in every case. Most people accept differences between humans and animals in many areas, from what we eat to whom we vote for, and those differences need not be arbitrary. What the objection does establish is that research must justify each use. That is why UK law requires every project to be licensed and expects researchers to replace animals where they can, reduce their number and refine methods to cause less harm. A licence is granted only when the expected benefit is judged to outweigh the harm to the animals, which is exactly the weighing this essay defends.

   NoteA partial concession, then a reasoned limit on it. The UK's licensing system and the three Rs are real and are used as evidence for the conditions in the first paragraph.

6. So the answer depends on the case, but not in a vague way. Research that could be done with cells, computer models or human volunteers should be; research that offers little benefit should not go ahead. Where neither is true, using animals to protect human and animal life is justified, and the duty is to cause as little suffering as possible. A society that accepts the benefits of medicine owes animals at least that much honesty about the cost.

   NoteThe conclusion turns "it depends" into a test the reader can apply, which is the difference between a balanced essay and a vague one.

What would make it stronger: Name one kind of research that would fail your test, such as testing cosmetics. A concrete no makes the conditional yes believable.