Someone at dinner says "well, the strong survive, that's just nature." A sentence like that needs no explanation; everyone nods.
But three judgements are packed inside it: that there exists a ladder running from low to high; that whoever sits lower deserves to be removed; and that the whole thing has a direction, and the direction is called progress. Not one of the three is in anything Darwin wrote.
A theory gets misread this badly not because readers are careless. Quite the opposite — the misread version is easier to follow, easier to say, easier to use. The trouble is that it is also wronger, and wrong with consequences.
You have almost certainly seen the picture: a stooping ape on the far left, walking rightwards, back straightening step by step, until it becomes an upright human. It is on T-shirts, in adverts, on textbook covers.
It has a definite source: a 1965 illustration in Time-Life's Early Man, by the painter Rudolph Zallinger, originally titled "The Road to Homo Sapiens". The original was a row of separate fossil reconstructions, laid out as a queue — and then the queue walked off on its own for sixty years.
The queue implies three things, none of which holds: that each figure is what the previous one turned into; that further right means higher; that the road has a destination and we happen to be standing on it. The real shape is a branching bush — several kinds of hominin were alive at the same time, cousins rather than parents and children, and most of the branches ended partway up.
Darwin himself detested the ranking. In an 1837 notebook he wrote that it is absurd to talk of one animal as higher than another. Around 1845 he pencilled a reminder to himself in the margin of his copy of Vestiges of the Natural History of Creation — a bestseller of the day arguing that evolution is one long climb — never to say higher or lower.
The hardest counterexamples are the organisms that went the other way. A tapeworm has no gut at all: it sits in its host's intestine and absorbs already-digested food straight through its body wall, and its ancestors had guts. Sacculina is a barnacle whose adult keeps no eyes, no limbs, not even a segmented body. Measured with a "steadily more complex" ruler, both are regressions; yet they come off the same machine as a giraffe's neck — variants turn up by chance, and the ones whose bearers leave more offspring stay. Sitting inside someone else's intestine, dropping an entire digestive system is a bargain.
But surely it is true that the most complex organisms have become steadily more complex? That one is still argued over. Gould offered an explanation: life started pressed against a wall, since complexity cannot fall below zero, so even if every step were random there is only one direction to spread in, the average drifts upward, and no "trend" is needed to push it. The argument is unsettled, but a rising average is not every lineage rising.
"A sunflower turns in order to catch the sun." "Trees make oxygen so that animals can breathe." Both sound unobjectionable. The second is false: oxygen is a waste product of photosynthesis and has nothing to do with whether animals need to breathe; Earth had oxygen long before anything breathed it.
Kelemen and colleagues ran an experiment in 2013: eighty working scientists — chemists, geoscientists and physicists from universities like Harvard, Brown and MIT — judged a set of statements true or false, with sentences of exactly that kind mixed in. Given unlimited time they called such statements true about fifteen percent of the time; put under time pressure, with only seconds per item, it rose to nearly thirty, while their accuracy on control items held. This is not a knowledge problem — explaining the world by purpose is a default setting, and training covers it rather than removing it.
Once "in order to" enters a sentence it smuggles in three things. Foresight: as if the variant knew it would be useful later. A beneficiary: "for the good of the species" is the most common version, and it is wrong — a trait is not kept because it helps the species; there is only one criterion, that the individuals carrying it leave more offspring who themselves go on to breed. And necessity: as if the thing had to happen, rather than happening not to die out.
Biologists use this shorthand daily, and that is fine, provided it can be cashed out on demand. Pittendrigh coined the word teleonomy in the 1950s — read it as "an imitation of purpose" — meaning that the appearance of purpose is an output of selection, not its cause. The test is concrete: rewrite the sentence in the form "this variant appeared, and the individuals carrying it left more offspring". If it rewrites, the shorthand is safe; if it does not, it is still an unredeemed promise.
The rewrite also forces you to name a place. "Dark is better" carries nothing; "in that stretch of soot-blackened woodland, dark moths were harder for birds to spot" carries something. Later the soot was cleaned up, the bark went pale again, and the dark form fell back — the same trait, filtered in opposite directions within a century.
Back to that sentence at dinner. "Survival of the fittest" was not written by Darwin. It was written by Herbert Spencer, an English philosopher, in Principles of Biology in 1864. Spencer's own account of evolution was published a few years before the Origin and turned on everything moving towards greater perfection — not the same thing at all. Darwin only took the phrase into the fifth edition of the Origin in 1869, on Wallace's urging, hoping to dodge the flavour of somebody doing the picking that clung to "natural selection". The trouble he got in exchange was larger.
Because in biology "fitness" is not an adjective, it is an account: in one specific environment, over one specific stretch of time, how many offspring capable of breeding the carriers of a trait left on average. That is all of it, with no sense of better, stronger or more advanced attached. Where malaria is common, carrying one copy of the sickle-cell variant gives better resistance to it, carrying two brings severe anaemia; move the same variant somewhere without malaria and only the cost remains. Fitness travels with the environment, and when the environment shifts the ranking is redrawn.
So the sentence "the strong survive, that's nature, therefore…" breaks at "therefore", whatever comes next. The break has a name in philosophy: Hume pointed out in the eighteenth century that no pile of statements about what is yields a statement about what ought to be, and Moore later named the move of equating "good" with some natural property the naturalistic fallacy.
This is not a matter of taste. In the twentieth century people did step across that break. In the United States, from the 1920s to the mid-1970s, more than sixty thousand people (some counts approach seventy thousand) were sterilised without their consent; in 1927 the Supreme Court upheld such laws eight to one in Buck v. Bell. Among those selected, poor people, ethnic minorities, deaf and blind people and people labelled "feeble-minded" were represented out of all proportion. The justification offered was improving the quality of the population, and the vocabulary was borrowed from biology.
The label "social Darwinism" is itself disputed: it was rarely used in the nineteenth century and became common only after Hofstadter's 1944 book, and historians have since pointed out that many of the people it covers — Spencer among them — did not rest their case on Darwin's theory at all. However that dispute settles, it leaves the previous paragraph untouched: borrowing biology's vocabulary to endorse a human arrangement is something that happened, and the cost was real.
One more thing has to be said flatly. The great majority of human genetic variation sits inside any one population, and the portion between populations is small. So "that population is more advanced" does not need a moral argument to defeat it; it fails at the level of the data. And "more advanced", as above, has no definition in this field at all.