
is the textbook case of rigid learning: a hatchling fixes on whatever it meets first and follows that. A handful of animals have shown , but only after long stretches of . So two researchers briefly showed newly hatched a pair of objects, either matching or mismatching in shape or color, and then offered pairs of completely new objects. Which pair did the ducklings follow?
They followed the relation, not the objects. Ducklings briefly exposed to a pair that was the same later preferred to follow new pairs that were also the same, and those exposed to a mismatched pair preferred new mismatched pairs. The shapes and the colors were different the second time, so only the abstract relation carried over. No reinforcement training was involved at any point.
Holding a logical relation between two stimuli, and then applying it to stimuli you have never seen, is called relational concept learning. It is abstract in the strict sense, because it survives changing everything concrete about the examples. A few species manage it, usually after extensive reinforcement training.
Imprinting sits at the opposite end of the reputational scale. A newly hatched bird meets something, follows it, and keeps following it, which looks like the most rigid and least intelligent learning in the catalogue. Putting the two together is the experiment.
Newly hatched domesticated mallards were briefly exposed to a pair of objects. For some ducklings the two objects matched in shape or in color; for others they differed. Later the ducklings met pairs of entirely new objects, and they preferred to follow the pairs that repeated the relation they had been exposed to.
Nothing concrete carried over, since the second pairs shared no shape or color with the first, so the only thing the duckling could have retained is the relation itself. The authors draw the conclusion plainly: even in a fast and seemingly rigid form of learning, the brain is working with abstract conceptual reasoning, a faculty usually reserved for animals we rate highly. Keep the scope narrow, though. One relation, one behavior, and following a pair of objects is a long way from reasoning in general.
wat. A duckling shown two matching objects went on to follow other matching pairs it had never seen, having been taught nothing and paid nothing.
wat wat
Abstract conceptual reasoning is normally credited to the animals we already think are clever, and reached only after heavy training. A duckling got there within hours of hatching and with nothing offered in return.
How do we know?
Martinho and Kacelnik briefly exposed newly hatched domesticated mallards to a pair of objects matching or differing in shape or color, then measured which novel pair the ducklings chose to follow.
What is still unknown
Following a pair of objects is one narrow behavior, and this tests one relation, sameness against difference. It does not show a duckling reasoning about anything else.
Learn something?
Where to next
You were told you are a visual learner. Should you be taught only in pictures?learning
Does living abroad really make a person more creative?learning
Does computer read-aloud help you understand a text, or only read it out?learning