Skip to content
NatureHQ

A guided journey · 5 stops

The untested obvious

Five things everybody knew about animals, and the five different ways they turned out to be wrong — or unchecked, which is not the same thing.

The route

  1. Constriction
  2. Mutualism
  3. Scavenging
  4. Convergent evolution
  5. Collective behaviour

Most of what this site corrects is not a myth. Nobody invented the idea that constricting snakes suffocate their prey, or that hyenas live off lions' kills, or that cleaner fish reduce parasites. Each was a reasonable inference from what could be seen, repeated by people with no reason to doubt it, and left standing because nobody had done the specific thing that would have tested it.

That is a more interesting failure than a myth, and it fails in identifiably different ways. This route walks five of them, and the order is by how the error was made rather than by subject. A plausible mechanism nobody measured. An effect everybody assumed. A reputation that was an artefact of when the observers were awake. A count with nothing to compare it against. And a simulation that looked so right it did the explaining for thirty years.

None of the animals here is the point. The point is that "obvious" and "established" are different words, and the gap between them is where most of this corpus lives.

  1. Stop 1 of 5. Reptiles and amphibians

    A mechanism nobody had measured

    Start with the cleanest case. Constricting snakes were said to suffocate their prey — coils tightening on each exhalation until the animal cannot breathe. It fits everything visible from outside, it was in textbooks for a century, and nobody had put a pressure line into the question. When somebody did, arterial pressure collapsed within seconds and the blood chemistry matched circulatory failure, not asphyxiation. The old account was not careless. It was untested, which is a different fault and a more common one.

    Read ConstrictionSuffocation was the textbook answer for a century, and had never been measured. When it was, circulation stopped in seconds — by a completely different route.

  2. Stop 2 of 5. Ecology

    An effect everybody assumed

    Now a subtler version, where nothing was even claimed incorrectly. Cleaner fish were observed picking parasites off client fish; parasites were known to be on the reef; the effect on parasite loads was assumed. Somebody removed cleaners from patches of reef and counted, and the loads did rise — so the assumption held. Included precisely because it held: the lesson is about what had been established, not about who was wrong. Reading it here also sets up the page’s harder point, that the cleaner would rather eat its client’s mucus.

    Read MutualismA cleaner fish would rather eat its client’s mucus than its parasites. Mutualism is not friendship — it is an arrangement that needs holding in place.

  3. Stop 3 of 5. Ecology

    A reputation made of working hours

    Third failure mode, and the one that generalises furthest. Hyenas were the scavengers and lions the hunters, and years of direct observation found the reverse: hyenas hunt most of what they eat, and lions take kills from them. The reason for the error is identifiable. Hyenas hunt at night and are seen standing at carcasses in the morning; lions are conspicuous at any hour. Anyone watching in daylight would have recorded exactly the reputation that stuck. It is not a claim about animals at all — it is a claim about sampling.

    Read ScavengingCarrion looks like free food. So why has no walking animal ever managed to live on it?

  4. Stop 4 of 5. Ecology

    A count with nothing to compare it against

    The fourth mode moves from observation to analysis. A widely reported result found convergent substitutions between echolocating bats and dolphins across the genome. The objection was specific: with twenty amino acids and a great many sites, any two lineages land on the same residue somewhere, so the number means nothing until it is set against non-echolocating pairs. Supply that comparison and the excess was no longer clear. This site does not adjudicate the dispute, and the reasoning behind it is the same one applied to every behavioural claim here.

    Read Convergent evolutionVertebrate and octopus eyes are the classic case of convergence — and what makes them convincing is that they are built differently.

  5. Stop 5 of 5. Behaviour and cognition

    A model that looked too right

    End with the hardest one, because it is the failure mode that feels least like a failure. Three local rules in a 1987 computer graphics model produce something indistinguishable from a flock, and they became the standard public explanation of murmurations. What that demonstrates is sufficiency — the appearance is reachable that way. Whether starlings do it was a separate question, and when somebody finally reconstructed real flocks in three dimensions, the birds turned out to track a fixed number of neighbours where the models had assumed a fixed distance. The model raised the question precisely enough to be answered, which is the best thing a model can do and is not the same as answering it.

    Read Collective behaviourA starling tracks about seven neighbours — a number, not a distance. That is what holds a flock together when a predator stretches it.

Where this leaves you

You end with five distinct failure modes and a working habit: when you meet a confident claim about an animal, ask what measurement it rests on and what would have happened if the claim were false.

The most uncomfortable stop is the last. A model that reproduces a phenomenon feels like an explanation, and it is a hypothesis with a demonstration attached — and in the flocking case the early models guessed one detail wrong, which the field measurement had to catch.

What the route does not settle: how much else in this corpus is currently in the same position, resting on something reasonable that nobody has yet done the specific test for. Some of it certainly is. The pages say so where the answer is known.

Last reviewed 2026-09-04. All guided journeys