Have you ever watched a child as they attempt to identify animals? Point at a dog, and maybe they will think it is a squirrel.

“That’s a squirrel, right? Or is that a chipmunk?”  a child may ask.

“No, that’s a dog. See, it looks like the dog that uncle has,” you might respond.

Why did the child decide that the dog was a squirrel? After all, dogs surely don’t look anything like squirrels, right?

As it turns out, squirrels do in fact look like dogs!

Wait a minute! You are telling me that a dog looks like a squirrel? I mean, don’t squirrels look a lot smaller than dogs? And who would even have a squirrel as a pet? I mean…….

As it turns out, they look similar when you view them through the lens of schemata, namely, when you look at it through the ‘components’ the way that a computer does.

Piaget’s Theories on Knowledge

Jean Piaget was a famous Swiss psychologist in the 20th century, whose work on child development is still important today. Piaget’s work led him to reconsider how the mind works, and a large part of his work was with the education of children.

Going back to the squirrel-dog example, Piaget might explain the child’s belief that the dog is a squirrel by explaining that the child looks at the dog as another four-legged furry animal with a tail, much like a squirrel. Looking at it this way, if a child has never been able to put a name to a dog but has learned what a squirrel is, it is only reasonable that the child views the dog as a squirrel.

When the child here views a dog, the dog fits the schema that the child has on what a squirrel is, namely a four-legged furry animal with a tail. Confronting a new object, the child can either choose to assimilate the dog into the schema as simply another squirrel, or they could accommodate the dog as a new object distinct from a squirrel. This process occurs whenever the child encounters a new piece of stimuli. 

Where else can this process be seen? In computers, of course. In machine learning and neural networks, computers have to learn to distinguish between different objects. One way which computers are trained is through repeated exposure to pictures of what say, a dog is. Through repeated exposure, a computer can pick up that dogs have a few common characteristics. 

The computer runs into a similar problem to the child when it encounters a picture of a dog that deviates from the norm. In that case, like the child, the computer will report that the picture of the dog is not a dog, when in fact, it is. For example, if dogs have been flashed with only a picture of their faces, then the computer might have a hard time recognizing a dog if you show a picture of the dog from the side, or from behind. In that case, the computer is unable to assimilate the dog into its schema, believing falsely that it is not a dog. 

Piaget may not have envisioned the connection between his work and the frontiers of computer science, but it is quite interesting to note the connections that are present. In the future, I believe that computer scientists and child psychologists should come together so that we may develop further advancements in computing, as well as child education.