blogjou
It seems like at least the European CORONA crisis is coming to an end, so I need another socially accepted excuse for never being around anywhere. A blog!The prevalence of insignificance makes postdocs perish like fruit flies
Significance arises only where a relationship is not due to chance. In these cases, the null hypothesis is refuted and an alternative hypothesis vindicated. Impotantly, the null hypothesis does not need to be proved; it is only refuted trough expriments demonstrating significant results. […] The fact that the overwhelming majority of results are insignificant proves the wisdom of the presumption of chance. The prevalence of insignificance also explains why postdocs end up perishing like fruit flies rather than publishing.
Scientists are supposed to presume the null hypothesis
The trouble is that biologists consider function rather than neutrality as the default state. They presume significance - selection - rather than chance, which in this case is mostly drift. This stance is unscientific. Scientists are supposed to presume the null hypothesis, which roughly postulates that every relationship is the result of chance unless proved otherwise. The burden of proving significance weighs on the scientist except in evolutionary biology, where significance is the starting point. In this field, the burden of proof lies with whoever argues that chances determines the emergence, morphology, and size of a particular trait.
Generality is the signal and uniqueness the noise
Fundamentally, neutraity helps us keep our attention on the case that matters most, which is the general case. Aristotle described the proper focus of science when he aserted that “the universal is more important than the perception of particular cases.” Yet biologists obsess over uniqueness, hence Romanes’s unheeded admonition that specific characters - traits that distinguish a species from its nearest relatives - are precious to taxonomists but may well be useless for the organisms themselves. The traits on which species boundaries are based need not be useful; the need only be nonlethal. Indeed, as I detail in Chapter 7, the most useful traits tend to be those shared by many species, those that have proven their value to survival and reproduction across geological time and thus have persisted through bottlenecks. Generality is the signal - the selected - and uniqueness the noise. For evidence, we might look to our own eyes and genomes and fingerprints. Every human iris has its special patterns of colors, every genome its unique contents, and every fingerprint its singular topography of ridges and valleys. These signatures are useful to criminal investigators and designers of biometric passwords; they are biological indicators of the individual and no one else. All useful traits resemble each other; each neutral trait is special in its own way.
Nineteen tails in a row may still be due to chance?
The textbook example of the null hypothesis is the flip of a coin. Each coin flipped has a fifty-fifty chance of landing on heads or tails. According to the lat of large numbers, one has to repeat the flip many times in order to achieve the expected 1:1 ratio of heads to tails. How many times? The consensual number, twenty, was theorized by the statistician an geneticist Ronald Fisher. According to Fisher’s significance level, nineteen tails in a row may still be due to chance, but twent signifies a loaded coin. In experimental science, results obtained twenty times are said to be signfiicant, refuting the null hypothesis.
Aversion to neutrality is also embedded in language
Aversion to neutrality is also embedded in language. In Enlish, neutrality is expressed through the cancellation of other states. One may be nonaligned, disinterested, detached, unbiased, asexual, indifferent, or impartial. In this way, neutrality is coded as exceptional rather than normal. The normal state is one of preference or tendency, and neutrality is its negation. Linguists would describe neutrality as marked, which is to say uncommon, while the absence of neutrality is unmarked.
A universal theory is dubiously derived from an extreme case
If archipelagos are not nature simplified but rather nature in exceptional form, then the Galápagos is the exception to beat all others. Although situated in an equatorial region, the Galápagos archipelago has an exceptionally harsh climate. It is a kind of tropical desert - nothing like lush mainland Ecuador. Darwin notes in his Journal of Researches, “This archipelago, instead of possessing ahumid climate and rank vegetation, cannot be considered oherwise than exremely arid.” In average years, only the highest altitudes of the larger islands receive enough rainfall to support tropical life. The littoral and inland areas of the islands are classified as “arid” and “very arid” and are covered by brown and gray vegetation. Rainfall also varies considerably with altitude, between islands, and over time. Such conditions foster competition, the backbone of natural selection. One millimeter of beak length might be a question of life and death on the Galápagos’s lava, but two meters of neck length are negligible in the African savanna. A universal theory is dubiously derived from such an extreme case.
The Galápagos archipelago is unique
The finch, it turns out, is a good model for the study of natural selection, but this does not mean that the evolution of its beak followed a typical, much less unbiquitous pathway.
No lone road to the promised land
Alongside the numerous Galápagos species that did not stimulate Darwin’s genius, even the finches demonstrate D’Arcy Thompson’s dictum, announced in this book’s epigraph, that there are many paths to survival. Some were traced by darwin and his acolytes, some by Romanes, some by other theorists. Perhaps some have not yet been imagined or discovered. Nature may follow one or another of these paths more frequently, but none is the lone road to the promised land.
Alternative pathway of evolutionary biology
So Mendel got lucky, too. Had he continued with mice, he would not have been able to understand the genetic basis of the experimental outcome. But genetics would have been the same because Mendel’s laws, being universal, would have been discovered by others. In fact, they were - thirty-five years after he first presented his findings. The same is probably not true of evolutionary biology. If Darwin had not stumbled onto the finches, evolutionary biology would likely have taken another course: one less selectionist and less prone to the fallacies of the domestication analogy and to capitalism.
Nature's wide ranges are counterproductive
Nowhere is this antinomy more flagrant than in physiological ranges. On the farm, ranges keep narrowing; in the wild, as I discuss in Chapter 6, they keep widening. Wide ranges are counterproductive because they drive specimens away from the optimal point. That is why optimizers of all stripes declare war on quantitative variability. Consider what has become of the domesticated cow. The size of its ancestors must have been as variable as the behavior of wild foxes and the look and flavor of uncultivated apples. But geneticists, zoologists, bankers, marketeers, engineers, and consumers took it upon themselves to fabricate, sell, and buy the most cost-efficient cow. Their success is evident in a remarkable degree of standardization. Less than a hundred pounds separate the heaviest and lightest breeds, from Herefords at 1,419 pounds to Gelbvieh at 1,323 pounds. Nature’s cattle don’t have to meet the efficiency targets of business; only artificial selection grudges pennies like that.