Field notes on things that run themselves
The Half That Wasn’t Listening
Watch a greylag goose retrieve an egg that has rolled out of its nest and you see one smooth, continuous motion: the neck extends, the underside of the bill curls against the egg, and the head draws back in a long, careful arc until the egg is home. Now have someone lift the egg away mid-motion, quickly and quietly. The goose keeps going. The neck still draws back on schedule, the head still tucks in along the same slow arc, cradling nothing at all. It isn’t confused, and it isn’t faking it. It finishes the motion exactly as if the egg were still there — because for half of what you just watched, the egg was never really the point.
Ethologists call this a fixed action pattern: a stereotyped, species-typical sequence of behavior that a specific trigger — a sign stimulus, or releaser — switches on as a whole, and that then runs to completion under its own power, indifferent to whatever happens next. The idea is older than the famous goose. Oskar Heinroth, a Berlin Zoo biologist who spent decades studying the instinctive behavior of ducks and geese, described these rigid, inherited movement sequences as early as 1910, arguing they held steady enough within a species to double as evidence for its family tree. His student Konrad Lorenz spent the next two decades turning that observation into a working theory of instinct.
The goose case became that theory’s defining proof during one especially productive stretch. Lorenz met the young Dutch biologist Niko Tinbergen in 1936; the following spring, Tinbergen spent weeks at Lorenz’s family home in Altenberg, outside Vienna, and the two of them tested instinct experimentally almost for the first time. In one strand of work they flew cardboard hawk silhouettes over young turkeys to see which shape read as predator. In another, they took the egg-retrieval apart. Their joint analysis, published in 1938, is still ethology’s founding case study.
An incubating greylag that spots a displaced egg rises, walks to it, and lowers its neck so the underside of its bill rests against the egg’s far side. From there, the retrieval is really two motions layered on top of each other. One pulls straight back toward the body, drawing the egg up and over the rim of the nest. The other works side to side, continuously nudging the egg back toward the goose’s own midline whenever it starts to wander off course — which, being round, it constantly does.
Lorenz and Tinbergen’s real finding wasn’t just that the whole sequence runs on rails. It’s that only one of those two motions does. The backward pull is the fixed action pattern proper: ballistic, ruled by nothing but its own internal timing, and the part that keeps going even with the egg gone, tracing the same arc through empty air. The side-to-side correction is a different kind of thing entirely — a taxis, a movement continuously steered by the actual, ongoing feel of the egg rolling against the bill. Take the egg away and that half vanishes on the spot, because there’s nothing left to correct against. What looks like one seamless action is a closed loop and an open loop running at the same time, stitched together so tightly that for years no one thought to ask whether they were even the same kind of thing.
The trigger for all of this is almost comically unfussy. Tinbergen found that a brooding goose offered a choice between its own egg and an oversized volleyball will roll in the volleyball first, and keep at it longer. Goose eggs, golf balls, doorknobs, a plaster model too large for any goose to have laid — anything round, convex, and roughly egg-scaled will do. The releasing mechanism isn’t reading “egg.” It’s reading a handful of crude features, and it will happily overreact to an exaggerated version of them: the same soft spot later named the supernormal stimulus.
That word fixed has taken a beating since 1938. In 1974 the ethologist Wolfgang Schleidt published a paper bluntly titled “How ‘Fixed’ Is the Fixed Action Pattern?”, opening a reappraisal that has never fully closed. Real fixed action patterns turn out to have a genuine, if narrow, range: they grow easier to trigger the longer they go unperformed, they can fire in a stronger or weaker version depending on the animal’s own state, and in the strangest documented cases — an isolated male dove courting an empty corner of its own cage — they can eventually fire with no releaser at all. Much of the field now prefers the more modest term modal action pattern, treating “fixed” as a description of the typical case rather than an iron law. The goose’s backward pull is still about as close to genuinely fixed as instinct gets. It just was never quite the whole story — the taxis half, arguing against pure rigidity, was sitting inside the original 1938 paper the entire time.
Nearly everything this publication has covered holds its shape by continuously answering to the world: a flame reading its own oxygen supply, a heartbeat reading its own stretch, a voice reading a room it is also filling. The goose’s egg-roll is the rare case built the other way. Triggered once, half of it stops listening the instant it starts. Watching it fail — a beak closing patiently around nothing, on schedule, on the correct arc, missing only the thing it was ever for — is about as close as instinct gets to showing you its own wiring diagram.
One loop I’m watching
Next: thousands of meters down, past any sunlight a plant could use, whole communities of tube worms and blind shrimp build a standing ecosystem fed by nothing but chemical energy leaking out of the Earth’s own interior — the same “so do you” thesis as a coral reef, with the sun swapped out for the planet’s own heat.
Tip: the ← and → arrow keys move between issues.
Prefer feeds? Follow by RSS or JSON Feed.
Want something playable between issues? Thought Toys has its own email list and RSS feed.
New to The Standing Wave? Start here →