Berkeley psychologist Alison Gopnik showed kids a box that played music. Kids turned it on by placing the right pair of blocks on top. For some kids, any two different blocks would turn on the music box and for other kids, any two same blocks would start the music. Gopnik showed them how it worked, demonstrating a pair of blocks (same or different) that made the box play music and also a pair of blocks that left the box silent. She did it again, placing correct and incorrect pairs of blocks onto the music box, which played or didn’t play music. Then she asked kids to help her activate the box, offering the choice between a pair of “same” and a pair of “different” blocks to turn on the toy.
Three-year-olds couldn’t do it – they didn’t recognize that the “same” or “different” relationship between blocks was what caused the music. But 18-to-30-month-olds did. Unlike 3-year-olds, who did no better than random guessing, the study’s youngest participants were able to point to blocks that would turn on the music box 78 percent of the time.Gopnik calls this, “evidence for a genuine decline in relational reasoning between 18 and 48 months of age.” A paper describing the study is currently in press at the journal Cognitive Science.
Of course, humans eventually regain the ability to see the relationships between things. You can probably imagine that if you saw that a pair of pyramids and a pair of cubes activated a music box, but a cube and a pyramid or a pyramid and a sphere did not, you’d get the point. But there is a U-shaped curve in this brain skill in which kids can do it, then they can’t, and eventually they can again.
This seems like ticky-tack technical stuff until you start to consider what it might mean. It might, in fact, be an illustration of how and when a child learns to think.
Gopnik suggests that one reason 2-year-olds may lose their ability to see that a relationship between things can cause something is that, at the same time, they are solidifying their understanding that individual things themselves cause other things. Touching a bubble makes it pop. Flipping a light switch makes it bright. Showing bacon to a dog makes him drool. The same is true of a dad. This understanding of cause-and-effect is a major underpinning of logic.
Only, it may be that kids start to “think” before they’re able to think about their thinking; it may be that kids become so blinkered by the idea that a thing causes another thing, that they can’t imagine the possibility that in this case it’s the relationship between things that leads to the effect. Which block makes the box play music? It must be one of them!
Younger kids, without this bludgeon of logical certainty, are able to open their minds to a kind of intuition that lets them understand (maybe without really understanding) that it’s the relationship between blocks that activates the music box. Because they don’t “think” about it, they don’t come up with incorrect theories.
At least in the unscientific sample of my two kids, that sounds about right: When they were very young, they just kind of let the world wash over them, pawns to the subconscious. Then at age 3, you could see them start to think, and they used this new skill to create pretty steadfast confidence in things that weren’t actually true. Finally at around age 5, they tamed this dark and powerful magic of “thinking” and started to be able to evaluate their confidence in things — they became able to know what they knew and also know what they didn’t know, allowing the existence of other possibilities.
Maybe Gopnik’s finding that kids lose and regain relational reasoning is due to the messy on-boarding of our system of logic?
While this question is being sorted, it seems like we might all be able to play with this at home. I suggest silverware. Tonight I’m going to show my kids a fork and a spoon, and then distribute a chocolate chip. Then I’m going to show them two spoons, which does not elicit a chocolate chip. I wonder how many trials it will take before they recognize that “different” silverware earns a treat?
