| >In the new study, two crows were trained to create embedded sequences by pecking at brackets of colors and shapes on a screen. When the crows pecked a correct sequence, a chime sounded and the birds were rewarded with birdseed pellets or mealworms. If they pecked an incorrect sequence, a buzzer blared and the screen went dark for two seconds before the training resumed. >After a few days, the crows learned to peck correct sequences using bracket combinations they hadn’t encountered before at rates significantly higher than chance, Dr. Liao said. They pecked correct patterns at around the same rate as U.S. children and outperformed monkeys from the 2020 study, she said. That was the study >“Our research suggests that recursion isn’t the sole difference between human and animal cognitive ability.” I don't think anyone seriously thought this >Dr. Chomsky said he wasn’t convinced the crow study or earlier work including Dr. Ferrigno’s monkey study demonstrated recursion. He said he believes the ability is innate, not learned. >Rules people use to understand grammar and math go far beyond a crow’s recall of a few sequential patterns, Dr. Chomsky said. “It’s easy to show that humans have the rule in their heads,” he said. “There’s no evidence that corvids have the rule.” Not sure if this is what he means but I also believe the crow's recursion here may not be the same as that of a human. It is possible that the crows are doing the recursion in software, so to speak, and this software is compiled into the crows brain differently from the human, which calls the recursion instruction directly, so to speak. Then the crow might not scale to more complex tasks. This said, I'm sure now that they have found this ability of the crow more complex studies will be held to understand the nature and extent of the crow's ability. It has opened a new area of investigation. |