I've got an article in today's New York Times about one of my perennial fascinations—musical hallucinations. One of the reasons that I find this condition so interesting is that it gives us a look under the neurological hood. Our brains do not simply take in objective impressions of the world. They are continually coming up with theories, and they test them against perceptions every moment of our waking lives. It would be impossible to test them against a complete picture of reality, because the world is simply too complex and ever-changing. Instead, the brain makes quick judgments on scraps of information, revising bad theories that don't make good predictions or using good theories as the basis for actions. Some scientists argue that musical hallucinations are evidence that our brains even make theories about music. When we hear stray sounds, we match them to tunes in our memory, in a sort of internal game of Name That Tune. Unfortunately, some people can't test their theories well enough, it seems, and so they wind up thinking a church choir is singing in the next room, when in fact there is only silence. There's one line of evidence that supports this explanation of musical hallucinations that I didn't have room in the article to explore. It turns out that some people have an analogous problem with their vision. They suffer from a condition known as Charles Bonnet syndrome, in which they have visual hallucinations. In some cases, the hallucinations are nothing but textures or wallpaper-like patterns. In other cases, people may see a row of people floating in front of them. Reginald King, the elderly gentleman who described his musical hallucinations to me, also suffers from Charles Bonnet syndrome. He told me about how he would see patterns on the ceiling, or sometimes a cat or a dog running across his bed. Victor Aziz, one of the scientists I interviewed for this story, has noticed that some other people also experience both visual and musical hallucinations, and doesn't think it's a coincidence. It's possible that regions of the brain that handle processing complex structures of both sound and sight can short-circuit in a similar way, producing similar hallucinations. And interestingly, brain scans of people with visual hallucinations are strikingly similar to those of people with musical hallucinations. In each case, the higher information-processing centers become active even when the regions that normally relay information from the senses are quiet. If we accept a theory of what we see, it's as real as the theories of what we hear.