“Surely You’re Joking, Mr. Feynman”: Adventures of a Curious Character

A week or two after, I was thinking a great deal about how the brain works compared to how a computing machine works—especially how information is stored. One of the interesting problems in this area is how memories are stored in the brain: You can get at them from so many directions compared to a machine—you don’t have to come directly with the correct address to the memory. If I want to get at the word “rent,” for example, I can be filling in a crossword puzzle, looking for a four-letter word that begins with r and ends in t; I can be thinking of types of income, or activities such as borrowing and lending; this in turn can lead to all sorts of other related memories or information. I was thinking about how to make an “imitating machine,” which would learn language as a child does: you would talk to the machine. But I couldn’t figure out how to store the stuff in an organized way so the machine could get it out for its own purposes.

When I went into the tank that week, and had my hallucination, I tried to think of very early memories. I kept saying to myself, “It’s gotta be earlier; it’s gotta be earlier”—I was never satisfied that the memories were early enough. When I got a very early memory—let’s say from my home town of Far Rockaway—then immediately would come a whole sequence of memories, all from the town of Far Rockaway. If I then would think of something from another city—Cedarhurst, or something—then a whole lot of stuff that was associated with Cedarhurst would come. And so I realized that things are stored according to the location where you had the experience.

I felt pretty good about this discovery, and came out of the tank, had a shower, got dressed, and so forth, and started driving to Hughes Aircraft to give my weekly lecture. It was therefore about forty-five minutes after I came out of the tank that I suddenly realized for the first time that I hadn’t the slightest idea of how memories are stored in the brain; all I had was a hallucination as to how memories are stored in the brain! What I had “discovered” had nothing to do with the way memories are stored in the brain; it had to do with the way I was playing games with myself.

In our numerous discussions about hallucinations on my earlier visits, I had been trying to explain to Lilly and others that the imagination that things are real does not represent true reality. If you see golden globes, or something, several times, and they talk to you during your hallucination and tell you they are another intelligence, it doesn’t mean they’re another intelligence; it just means that you have had this particular hallucination. So here I had this tremendous feeling of discovering how memories are stored, and it’s surprising that it took forty-five minutes before I realized the error that I had been trying to explain to everyone else.

One of the questions I thought about was whether hallucinations, like dreams, are influenced by what you already have in your mind—from other experiences during the day or before, or from things you are expecting to see. The reason, I believe, that I had an out-of-body experience was that we were discussing out-of-body experiences just before I went into the tank. And the reason I had a hallucination about how memories are stored in the brain was, I think, that I had been thinking about that problem all week.

I had considerable discussion with the various people there about the reality of experiences. They argued that something is considered real, in experimental science, if the experience can be reproduced. Thus when many people see golden globes that talk to them, time after time, the globes must be real. My claim was that in such situations there was a bit of discussion previous to going into the tank about the golden globes, so when the person hallucinating, with his mind already thinking about golden globes when he went into the tank, sees some approximation of the globes—maybe they’re blue, or something—he thinks he’s reproducing the experience. I felt that I could understand the difference between the type of agreement among people whose minds are set to agree, and the kind of agreement that you get in experimental work. It’s rather amusing that it’s so easy to tell the difference-but so hard to define it!