Animal Experimentation: Is it Science?
By Nick Drake


This article looks at the use of animals in experimental procedures and attempts to understand whether or not such work is scientific, i.e. that experimenting upon live animals falls within accepted definitions of good scientific practice. The term 'animal experimentation' is used in preference to 'vivisection' although the two are often used synonymously: vivisection can also include experimentation upon healthy people in order to try to study human disease; but this controversial topic is not dealt with here.
Furthermore, no attempt is made to discuss a wide range of specific examples of experiments. Many individuals and organisations have written about such work many times before and it is available to the public. Instead, basic principles of the scientific method are introduced with the aim of answering the question: 'Are animal experiments scientific?'
What is science?
A concise definition of science is 'an organised and systematic activity that gathers knowledge about the world and condenses that knowledge into testable laws and principles'. The eminent biologist, E. O. Wilson, defines five diagnostic steps which can distinguish science from pseudo-science:
Repeatability:
The same phenomenon is sought again, preferably by independent investigation, and the interpretation given to it is confirmed or disgarded by means of novel analysis and experimentation.
Economy:
Information is abstracted into the simplest, most aesthetically pleasing form - scientific abstractions are 'elegant'.
Mensuration:
Generalisations can be made unambiguous if something can be measured properly using universally-accepted scales.
Heuristics:
The best science stimulates further discovery, often in unpredictable new directions, and new knowledge provides an additional test of the original principles that led to its discovery.
Consilience:
The explanations of different phenomena most likely to survive are those that can be connected and proved consistent with one another.
Does animal experimentation follow these principles?
Scientists design experiments to investigate particular questions. This is not possible in animal experimentation --- the animal comes as a whole, living system. The scientific method is best applied by varying one parameter in an experiment, ensuring everything else remains constant, and observing how that single change affects the overall system. Animals are complex systems: we do not know how to isolate specific properties in a living creature whilst leaving everything else unchanged.
Animal Experimentation and the Scientific Method
Animal experimentation is not scientific. It does not adhere to the five diagnostic steps listed above. There is no experimental model for the human species: All species and even individuals within a species differ from each other, and there is no known way to extrapolate accurately, reliably and repeatedly experimental observations from one species to another.
Experiments on whole, living animals (including people) are generally not reproducible. For example, the age of the animal can produce different, unpredictable results from the same stimulus. The time of day or year, and the conditions in which the animals are kept, can change the results of even simple tests. Such results are intrinsically false within the species concerned - it would be meaningless to extrapolate them to another species.
Perhaps even more alarming is the choice of animal for each experiment. The Home Office states that approximately 2.5 million animals were used in scientific procedures in 1998 (and it's estimated that at least another 4 million were killed as 'wastage', bred but never used in any experiment). The vast majority of these were rats and mice. But rats and mice are not chosen because they are the species which have the most similar biochemical reactions to those of people, but rather because they are cheap and quick to breed. Scientifically, one would expect experiments motivated by human illness to be performed on people with such illnesses; the second best choice would be healthy human volunteers; and if no person is available to act the guinea-pig, chimpanzees and other primates are the least worst animal choice. But chimps are expensive to keep, difficult to breed and can deliver a much worse bite than any mouse. Science is not a major factor when choosing the subjects of live animal experimentation. Animal experiments are not economical in the scientific sense.
The animal experimentation 'method' works like this: A researcher tries to model a human illness in a different species, perhaps by inoculating a pathogenic agent into an animal. This creates an infectious disease in the animal, different from the human version, which the researcher tries to cure with a variety of drugs. In other words, this method introduces errors at each step, which multiply geometrically. The probability that any cure found for the 'animal model' will work for humans with acceptably minor side effects falls closer to zero with every action the researcher takes. Moreover, there are no 'universally-accepted scales' to allow measurements of one species' reactions to be applied to another's. Generalisations from animal experiments are never unambiguous.
A case may be made, however, for animal experimentation to be labelled heuristic, since there are clear examples of serendipitous discovery. But many of these discoveries are observations of unexpected results: a drug having the opposite effect of that predicted, for example. Such discoveries almost always contradict original principles or hypotheses; the work is not truly heuristic.
Modern laboratories attempt to make experiments reproducible by using animals which are as standard as possible. Animals are fed standardized diets and kept in as identical conditions as possible. The result is not standardized animals, however, but abnormal, even sick animals, whose natural impulses have been paralyzed by such standardisation. It would be profoundly inaccurate to draw analogies between such animals and human beings, most of whom manage to eat varied diets, touch everything around them and face daily exposure to innumerable, random substances.
Biomedical research methods
It is important to realise that there are no 'alternatives' to animal experimentation, since such methods would be seen as equally valid and hence of no scientific merit. There are, however, many scientific biomedical research methods:
Cell and tissue culture techniques have many applications in research and great potential for replacing many unsatisfactory animal procedures. For example, substances intended for human use must undergo toxicity testing on human cell and tissue cultures. At present, much toxicity work is still carried out on whole, live (non-human) animals!
Epidemiology - studying diseases within whole populations can lead to huge gains in knowledge with high accuracy and low risk. Major advances in our understanding of, cancer and HIV, for example could be attained simply by observing their characteristics among specific population groups. Instead, vast sums of money are currently expended on laboratory research into cures for such afflictions, before we have any deep understanding of their propagation through society.
DNA profiling is certainly controversial, but will undoubtedly become a significant tool for pharmaceutical companies. In the future it is likely that prescriptions will be tailored to each patient based upon a DNA analysis of their probable responses to different drugs and, in principle at least, this could significantly cut back both animal and human testing programmes.
Computer modelling allows the researcher to explore many more possible combinations of potential medication than any laboratory programme can offer.
In summary, the natural models for human illnesses are sick people, not healthy animals or purpose-bred animals made sick to mimic human illnesses. Sadly, there are many ill people in the world and many illnesses for which we still need better prevention and/or cures. This is a situation which will not improve while governments, universities and pharmaceutical companies continue to mis-apply research efforts to animal experimentation. Researchers need to seek out the ill, wherever they are globally, and engage with them in the struggle to improve human health.
Conclusion
Animal experimentation is bad, lazy pseudoscience which can be replaced with numerous, scientifically rigorous, methods, such as clinical observation, epidemiology, cell and tissue culture studies, and mathematical models. Results obtained from animal studies and applied to humans can only ever be interpreted with hindsight. We can never know a priori if a substance or technique which 'works' in one species will work in another.
Animal experimentation is scientifically illogical and produces results which cannot be reproduced freely. It does not follow the standard scientific method. Medicine should not progress by trial and error because the 'errors' are human and animal lives.
Interesting Quotes
'No experimenter on animals can provide a single useful fact about human disease.' (D. A. Long, 1954, British National Institute for Medical Research)

'The idea, as I understand it, is that fundamental truths are revealed in laboratory experimentation on lower animals and then applied to the problems of the sick patient. Having been myself trained as a physiologist, I feel in a way competent to assess such a claim. It is plain nonsense.' (G. Pickering, 1964, Oxford University)

'It was by good luck that in the initial toxicity tests we used mice, because if we had used guinea-pigs we would have concluded that penicillin is toxic.' (Statement by Sir Howard Florey, joint Nobel Prize winner with Fleming and Chain for the discovery of penicillin)
References:
Croce P., 1999, Vivisection or Science? An investigation into testing drugs and safeguarding health. Zed Books, London.

The author, Nick Drake, has a background of research in the physical sciences and has been taught scientific methodology and processes at various UK universities.

Why is animal experimentation still tolerated in a sophisticated, civilised world? Perhaps because such basic truths as its non-scientific nature rarely reach the majority of people. So spread the word: Verifiable truth is of the greatest power, and always wins eventually. (sf)