Yesterday afternoon I went to a birthday party, where I played  hide-and-seek with the hosts' four-year-old little girl, L. It reminded me of the work of my friend Greg, a cognitive scientist who has used hide-and-seek as a task in some of his case studies [PDF]. (Long-time readers of this blog know that I go meta at the slightest provocation, even in very ordinary situations.)

 Why aren't young children very good at playing hide-and-seek? There are several reasons. For one, young children don't yet understand some basic spatial relationships. Have you ever seen children pour water betwen cups of different sizes in the bathtub? They're learning about relationships between shape and volume, and it's not always obvious to them that a tall thin container might hold less than a short fat one. Concealment means hiding behind or under or inside of something, concepts that are still a bit mysterious. More importantly, young children don't yet understand how to take someone else's perspective. In some situations, they'll automatically assume that what they see or know is the same as what others see or know.

Greg gives a nice example that will probably be familiar to most parents: In the first game of hide-and-seek, his "subject" (a three-year-old girl) found a hiding place by going into the next room and closing her eyes. (I hope the perspective-taking problem is obvious here.) My friend L. has gone beyond this stage but not very far: Before I finished counting to ten, facing a tree in the backyard, she shouted out, "I'm ready! Come find me!" while standing next to the nearest shrub. She continued to offer helpful advice ("I'm not there!") as I searched other nearby hiding places.

Like Greg, I decided to offer some strategic advice. "Hide somewhere I can't see you." This worked a bit better: She moved farther away, though still remaining in sight. Eventually L. chose a better hiding place, between two large bushes, where I couldn't see her when I finished counting by the tree. She still has some progress to make, though; it was only a matter of walking across the yard to find her. I suspect that L. hasn't yet figured out that you need a large, opaque object between two people for one to be concealed from the other. (Greg gives another funny example of his subject hiding underneath a glass coffee table.) I also suspect that L. may believe that a hiding place is good if she can't be seen from It's initial location, without considering that It will move later. I'm not worried--she'll figure it all out pretty quickly with experience.

So we had fun, the little girl playing hide-and-seek, the bigger guy playing hide-and-seek while thinking about cognitive development. And then we ate cake.