December 11, 2017

So far, so good

A leading article for The Ecologist Vol. 1 No. 16, October 1971. Republished in The Doomsday Funbook (Jon Carpenter Books, February 2006).

See ordering information for the Funbook.


The experts assure us that current levels of lead in our air and water are safe. Scientific endeavour, it would appear, is something that confers on a proposition some measure of credibility – perhaps even downright certainty. If so, how is this achieved? The answer appears to be: on the basis of observations made during laboratory experiments. A little reflection will reveal that such observations cannot provide any sort of certainty, and this for a number of reasons.

Firstly, sensitivity to different pollutants will vary with every individual. Thus, to determine with certainty what are the effects of a pollutant, such as lead, on human health, everybody would have to be tested. As this is not remotely feasible, a first compromise is required: we must content ourselves with testing a sample of people, and we must assume that the sample is representative. This may be probable, but it cannot be certain.

However, people do not on the whole volunteer to be the object of laboratory experiments. This renders necessary a second compromise: non-human animals must be used instead of people; and we must make a second assumption; that people will be affected by a particular pollutant in the same way as are those animals:- an assumption that forces us to make do with a still lower degree of probability. Of course probability could be increased by using animals as much like ourselves as possible, such as chimpanzees, but these are expensive and difficult to obtain. Hence there must be yet another compromise: experiments must be carried out on easily available mammals such as mice, still another compromise.

However, even mice cannot always be used. For in order to test the carcinogenic effect of a chemical, for instance, we would have to take into account that no more than one out of 500 people dies of cancer every year. If we regarded a 1 percent increase of the cancer rate as tolerable, the acceptability of the chemical concerned could only be determined by testing it on a sample of 50,000 mice. This might be conceivable if only a few chemicals had to be tested; but the chemical industry is pouring approximately 1,000 new chemicals into our environment each year, and already something like 50,000 are in use, the testing of which would require tens of millions of mice!

Imagine the environmental impact of harbouring such a large number of mice in our laboratories. The sewage bill alone would be prohibitive. There is another reason why mice may not do: the biological effect of pollutants is not immediately apparent. Cancer usually develops in a human some 20 to 40 years after exposure to a carcinogenic agent. Thus such effects can only be determined over a number of generations of laboratory animals.

Dr. Bryn Bridges of the Medical Research Council Cell Mutation Unit, Sussex University, has found that hamster cells cultivated in vitro best satisfy these requirements. To argue from hamster cells to people, however, must involve a still further compromise. In addition, we have been assuming that chemicals and other pollutants will have the same effect on us in different combinations as when affecting us separately, so ignoring the very important synergic effects.

Thus the effect of cadmium is much greater on people suffering from a zinc deficiency. The incidence of lung cancer among uranium miners who smoke has proved to be eight times higher than among non-smoking uranium miners. However, to test every possible combination and permutation of 50,000 different chemicals would require an astronomic number of white mice or a still greater number of hamster cells, at a totally prohibitive cost. Thus, regardless of the organism on which we choose to carry out our experiments, we are forced to make yet a further compromise: we must be satisfied with a very small sample of all the possible combinations, or, as the authorities do in the UK, simply ignore synergic effects.

In addition, we are still assuming that the correlations observed in laboratories constitute bona fide cause-and-effect relationships. Thus, when we observe the presence of a pollutant and a particular biological effect, we assume that the former causes the latter, but it could, conceivably, be the other way round. Or both could be the result of a third factor, or a variety of other factors. Empirical correlations cannot by themselves constitute evidence, and it is about time our experts realised this.

Four thousand people died from the London smog of 1954 before the Clean Air Act was passed. What sort of disasters are required to provide us with the ‘evidence’ that current pollution levels are getting out of control? The fact that scientists are not at the moment being called upon to count the corpses is the only grounds they have for stating that present pollution levels are acceptable. Indeed all that is implied by such statements is – so far, so good.

·Ω·

  • Twitter
  • Facebook
  • Digg
  • Reddit
  • StumbleUpon
  • Diaspora
  • Identi.ca
  • email
  • Add to favorites
Back to top