People submitted a lot of good ideas in response to my post about the talking Elmo book. Here is what I think the answer is, on the basis of some tests we did in the department.
One colleague studied the book for a little while, and said, “It has to be either optical, magnetic or ferroelectric.” Ferroelectric hadn’t occurred to me, but magnetic was my first thought. Moving that pen along that arrow seemed to me like swiping a credit card. But the problem with that, it seemed to me, was that magnets have only north and south poles (last I checked), while on any given page, I could get four different responses out of the pen (one of which was null). Also working against that idea was that you couldn’t use the pen on the back side of a page to get the same effect on the front side. You can hold pages up to the light and see through them, so there didn’t seem to be anything there that might shield a magnetic field from going through the page.
Optical thus seemed like the right choice, but whatever was happening wasn’t in the visible spectrum, as different colors could give the same effect. Another colleague who was playing with the book got a little lucky when he (by accident, as far as I could tell), covered a portion of the end of the pen with a piece of paper, and got a different response from when it was uncovered. Perhaps there was a sensor inside that responded to different amount of light that were reflected into it.
We went to his lab and examined the pen with an infrared viewer. Sure enough, when the end of the pen is pressed down, it emits light in the infrared. This is not my line of work, but I’m told that IR LED’s are quite inexpensive, so that makes them a good candidate part of the pen. We then examined the book under the microscope. My eyes aren’t really that great for these things, but my colleague insisted that he saw micropatterns that were different enough in different regions of the page such that they might have different reflectivities. More convincing was that he examined the patterns on two different pages, and predicted which pairs of colors on each page would gave the same response. He got them all right, which could have been luck, but probably wasn’t. And that pretty much sealed it for me.
