The art of scientifically missing the point

[RascalPuff]

And the Marooned Value of Null Experiments & Observations
Alexander Fleming inadvertently noticed the absence of proliferation in a colony of bacteria confined to a Petri dish, the neutrally sterile substance of which had been exposed to a specific strain of bread mold. He had anticipated proliferation of other microbes, but no such colonization emerged. It was a null experiment: resulting in the discovery and eventual mass production of penicillin – the stuff in the bread mold that prevented any other stuff from progenesis.

Michelson and Morley designed and built an interferometer to measure the ‘upstream’ speed of light cast from earth, and the ‘downstream speed of light cast from the earth. Their instrumentation was fairly tested, finely tuned and calibrated to measure any difference. There wasn’t any. Until further notice, it was the most important null experiment ever conducted in science. It led to Einstein’s Special Theory of Relativity in 1906.

Yet still, null experiments are generally considered to be of null importance and most frequently cast aside without further thought or contemplation. – K. B. Robertson (aka, Kai)

***************************

“Myths of Gender, by Anne Fausto-Sterling: is a fine contribution to the empirical literature on human gender differences. If her argument is correct, then this fundamental sociobiological theory for human universals has failed, leaving very little of substance for the putative revolution. Myths of Gender is a courageous book, for it centers its conclusion on the most undervalued and systematically ignored of all data – so called null results.

Few observers outside science (and not nearly enough researchers inside) recognize the severe effects of biased reporting (Refer, On the Structure of Scientific Revolutions, Kuhn). The problem is particularly acute, almost perverse, when scientists construct experiments to test for an expected effect. Confirmations are joyfully reported; negative results are usually begrudgingly admitted. But null results – the failure to find any effect in any direction – are usually viewed as an experiment gone awry… Most scientists probably don’t report such results at all – who has time to write up ambiguous and unexciting data? I call such non-reporting perverse because we cannot gauge its depth and extent Therefore we do not know the proper relative frequencies of most effects – a monumental problem in sciences of natural history, where nearly all theoretical claims are arguments about relative frequencies, not statements about exclusivity.

Over and over again in my career I have bashed my head against this wall of non-reporting. When Niles Eldridge and I proposed the theory of punctuated equilibrium in evolution we did so to grant stasis in phylogenetic lineages the status of ‘worth reporting’ – for stasis had previously been ignored as non-evidence of non-revolution though all paleontologists knew its high relative frequency. The critique of adaptionism fights a similar silence. A former student of mine recently completed a study proving that color patterns of certain clam shells did not have the adaptive significance usually claimed. A leading journal rejected her paper with the comment: Why would you want to publish such non-results?

This difficult and ostensibly unrewarding task forms the core for Fausto-Sterling’s book on claims for differences in styles of intelligence and hormonally mediated disparities in behavior between men and women. Over and over again she finds a disturbing pattern in the literature on “standard” gender differences. When differences are detected, they are usually in the direction proclaimed (though small in effect), but the great majority of studies report no difference and these have dropped from sight.

The study of gender differences suffers the same disabling bias. Measured differences are prominently reported, usually with much fanfare and much attention from the press. We really have no idea how often such differences are not found, for we don’t know what results are simply not published. But, at least we might tabulate assiduously all reported studies, positive, negative and null results.

Moreover, measured differences often correlate well with immediate cultural distinctions, leaving little space for the sociobiology of genetically grounded adaptation. For example, poorer spatial perception is frequently cited as a different cognitive style for women. Fausto-Sterling shows that such differences occur in societies that greatly restrict the autonomy and physical movement of girls during early upbringing, but have not been detected in cultures (including Eskimos) that grant equal freedom to boys and girls. I need hardly say that this pattern, if as general as Fausto-Sterling suspects, would remove any empirical reason for invoking socio-biological explanations for the central issue in the study of human universals. Few general theories can survive the collapse of their crucial case. – p. 37, 38 & 39, An Urchin in the Storm ('Cardboard Darwinism'), by Stephen Jay Gould

********************************

THE PEACE THAT PASSETH UNDERSTANDING:
The Mystery of Peace
(Excerpts from Chapter One:
The Cause of War, by Geoffrey Blainey)

For every thousand pages published on the causes of wars there is less than one page directly on the causes of peace. And yet, the causes of war and peace, logically, should dove-tail into one another. A weak explanation of why Europe was at peace will lead to a weak explanation of why Europe was at war. A valid diagnosis of war will be reflected in a valid diagnoses of peace.

One obstacle to studying international peace is perhaps the widespread assumption that it is perhaps the normal state of affairs. The assumption however is inaccurate, the talented American sociologist, Pitirim Sorokin, once busied himself by counting the number of years that some of the European countries spent at war. He found that Russia, the land of his birth, had experienced only one peaceful quarter of a century in the previous thousand years; in every other period of twenty five years she engaged in at least one foreign war. Since the year A.D. 901, he estimated, Russia had been at war in 46 of every hundred years… He found that England, since the time of William the Conqueror, had been engaged in at least one foreign war somewhere in Europe of the tropics for 56 of each hundred years. Spain experienced even more years of war.

There is another reason for the lack of detailed analysis of the causes of peace. For historians it is a powerful reason. They are usually shackled to the available evidence; in studying events they depend heavily on diplomatic documents, memoirs, newspapers, pamphlets and other written or printed records which have been handed down. These exist only because some politician, soldier or spectator possesses some opinion or news which he wished to communicate. Each record was as oriented to its author’s sense of news as a daily newspaper is oriented in the twentieth century. Since war was more newsworthy than peace, the records ostensibly said far more about the causes of war than peace. And yet the records of peace are extensive as those of war. They are simply less obvious. Any faded document which illuminates the causes of war must by implication also illuminate the causes of peace. Any document which discusses an international crisis that ended peacefully is a mirror of both peace and war. Nevertheless it is easy to see why peace often appears to be a newsless vacuum, a limbo for which scant explanation is necessary.

Historians’ explanations of peace in modern times are centered on the nineteenth century. Two long periods in that century were remarkably peaceful. One ran from the Battle of Waterloo to the short wars of 1848 or to the Crimean War of 1853. The other period of peace ran from the end of the Franco-Prussian War in 1871 to the close of the century, though the most common opinion is to assign the end of the long peace to 1914. Each era of peace therefore ran for about one generation. It is perhaps significant that while each war in history is given a name, no matter how short its duration or how slight its consequences, these long periods of peace have no accepted name.

These peaceful periods were not devoid of war. Their wars – in contrast to those of other periods – were simply fewer and shorter and were rarely between major powers. In the eyes of most contemporary periods however these periods were unusually peaceful, and they inspired confident predictions that a millennium of international peace would ultimately prevail…

To be continued...

[RascalPuff]

THE STATISTICS OF CONFLICT & RESOLUTION

To be continued...