602 Nature Vol. 293 22 October 1981 CORRESPONDENCE If Dr Pearson wishes to dispute the general then used to compute the reading ease of the Ethics for animals applicability of natural selection he should passage (RE) from the equation: StR- As noted in your editorial (Nature 17 probably begin by attacking the ubiquity of RE = 206.835 -0.846wl-1.015sl September, p.173), human research is already these conditions in nature. I think that it is controlled by ethics committees, and so we safe to say, however, that he will not meet Flesch produced a classification of RE which have some experience that can predict the with much success. is summarized in Table 1. We have chosen effects of introducing their non-human ROBERT P. GENDRON every twentieth letter published in Nature from counterparts. Duke University 1930, at ten-year intervals, to 1980 and An ethics committee that takes itself Durham, North Carolina, USA calculated RE for the first hundred words in seriously is a major obstacle to a research each. The first hundred were chosen in the worker. By the nature of things, its members belief that here the authors should be writing are senior, having their own productive Gold solution most clearly, attempting to explain the research days well behind them. The SIR- Your editorial on returning to the gold purpose of their work and its significance to supplicant is called upon to justify his standard (Nature 24 September, p.246) the general reader. Choosing the beginning proposition in advance of the evidence, so the showed an amazing degree of understanding rather than any other part also reduced the more original "look-see" type of work is of the topic. Gold has been a monetary chance of encountering complex formulae and discouraged. In arguing a nice point of logic, standard primarily because almost all that equations which do not readily lend themselves the committee is both cross-examining counsel has been mined in the past several thousand toRE calculations. Our results are shown in and judge. Research workers have many years is still around in a readily available Table 2. possible deadlines to meet, and delay imposed form. Mining at full tilt can scarcely make a It can be seen that there is a sharp drop in by waiting for one or maybe more meetings ripple in this huge, widely distributed bulk and RE when 1950 is compared with 1940. The may give the lead to other workers or cause an the major commercial uses of gold do not Student t-test shows this to be highly intending financial sponsor to take his grant destroy its availability, so that the supply is significant (P<O.Ol). TheRE changed little elsewhere. essentially independent of human agency. over the period 1950-80. Thus, our subjective But, having said all that, the ethics Your suggestion of the establishment of a impression that Nature was much easier to committee is a necessary evil, and the short-lived isotope as an alternative monetary understand during the 1960s than today does adversarial system sharpens the scientist's standard would never have occurred to me. not appear to be true, at least as measured by mind wonderfully. Since non-human animals Your observation that gold is unnecessary is reading ease. The significant drop in RE cannot be consulted or give consent, their correct. All that is needed is pure reason and between 1940 and 1950 was accompanied by "rights" must be safeguarded even more self-restraint and the world economy will run an increase in the number of letters published carefully than those of humans. I have never quite smoothly. Better yet, there will be no in Nature but, interestingly, no similar met a vivisectionist who approached his work wars. reduction took place in 1960 when the number thoughtlessly, yet standards of what is It is, however, not strictly correct that the of letters published was greater than ever acceptable vary enormously, and if nothing gold standard would force a country to before. else we should each be exposed to other balance its yearly budget. Borrowing of funds people's opinions when planning animal is still permitted under a gold standard. The Table 2 Reading ease (RE) of letters experiments. difference is that such debts would eventually published in Nature over half a century D.L.J. FREED have to be repaid rather than being inflated Year No. of/etters RE (s.e.) Research Institute of Scripps Clinic, out of existence. 1930 8 32.1 (6.3) La Jolla, California, USA DAVID DUNTHORN 1940 10 29.1 (2.6) CFSystems, 1950 29 16.5 (2.4) Oak Ridge, Tennessee, USA 1960 76 18.0 (1.5) Survivalist hopes 1970 59 15.0(1.6) 1980 41 12.8 (2.1) SiR- Barrie Pearson (Nature 3 September, A good read p.6) objects to Antony Flew's (Nature 16 July, SiR - That science is more complex that it It has been suggested 2•3 that, for a paper to p.192) "formulation of the principle of was fifty years ago is reflected in the be impressive, it must be almost or, even natural selection as non-random survival" on fragmentation of subjects into more and more better, completely unintelligible. For example the grounds that it is an "empirically empty specialities, each requiring at least one journal it has been demonstrated4 that, in the field of concept if the range of possibilities the non­ of its own to lend an air of respectability. The management studies, a journal's prestige random survivors were chosen from is hard-pressed scientist wishing to keep a increased as its readability decreased. unknown". Thus, he would have us believe tenuous grasp on the broad advance of science Although we are at present unable to test this that without such absolute knowledge turns, as he always has done, to Nature, idea rigorously for science journals we have biologists, as "creatures of reason", must Science and the like. But does the writing in used Current Contents to identify a dozen accept the possibility of random survival. This these non-specialist journals now also mirror letters published in Nature since 1950 which argument, however, is simply another one of the growing complexity of science? We have have been cited more than one hundred times. those "philosophical muddles" to which Dr applied an objective method of testing The mean RE of these letters was 25.0 with a Flew referred that continue to obscure the readability to letters published in Nature over standard error of 4.2 and perhaps the most creationist debate. the past fifty years and we find a highly celebrated of these, on the helical structure of The "range of possibilities" on which significant decrease between 1940 and 1950. DNA by Watson and Crick 5, had an RE of natural selection acts is ultimately dependent The readability after this date appears to be 51.9. These findings suggest at least that letters upon randomly occurring mutations. Many of stable. appearing in Nature need not necessarily be 1 these are deleterious, often with fatal The test, developed by Flesch , uses the cloaked in obscure and difficult language to be consequences. It is not necessary to identify number of syllables, words and sentences in a accepted for publication or to make a every possible mutation, its effects and its chosen passage to calculate the average word significant impact on the scientific world. frequency of occurrence, therefore, in order to length (wl) and sentence length (sl). These are eThe RE of this letter is 40.0. conclude that some allelic variants are much D. B. JACK less likely to persist than others. Table 1 Flesch classification of reading M. R. GREGG Of course, Darwin's theory of natural ease (RE) Department of Therapeutics and Clinical selection explains not only the elimination of Pharmacology, RE Style Typical example deleterious mutations, but also the spread of 0-30 Very difficult Scientific The Medical School, favourable ones. The generality of this theory 30-50 Difficult Academic University of Birmingham, UK lies in the generality of certain necessary 50-60 Fairly difficult Quality I. Flesch, R. J. appl. Psycho/. 32, 221-233 (1948). conditions. (I) Many more individuals are 60-70 Standard Digest 2. Siegfried, J.J. J. polit. Econ. 78, 1378-1379 (1970). produced than can possibly survive; (2) among 70-80 Fairly easy Slick fiction 3. Armstrong, J.S. Chemtech II, 262-264. 80-90 Easy Pulp fiction 4. Loveland, J. A cad. Manag. J/16, 522-524 (1973). these individuals there is heritable variation in 5. Watson, J.D. & Crick, F.H.C. Nature, 171, 737-738 90-100 Very easy characteristics affecting reproductive success. Comics (1953). 0028-0836181 /430602·01$01.00 © 1981 Macmillan Journals Ltd .