Bazelon writes that Schulz's research on early childhood learning indicates that "Most of the time, when children have a choice between a toy they've already played with and a new toy, they pick the new one. But if the toy they've already played with doesn't work the way they expect it to, they 'break their novelty preference,' . . . and keep playing with it—evidence that they're intrigued and exploring."

I don't understand what "breakiing the novelty preference" means here. Wouldn't the better interpretation be that children always keep their novelty preference? When an old toy is modified to work in a new way it essentially becomes a new toy from the perspective of novelty. There is no incentive to switch to a truly new toy until the novelty of the modified old toy wears off. It doesn't seem as if the novelty preference is ever broken; rather, it is always maintained if you accept a more general (and I think more accurate) definition of the novelty preference.