The editors have given the authors a difficult task: Pick out one of our “most important contributions to research” and write about it. Just the first part had me stumped: Most important in what way? I decided to subvert the question into telling about my most-cited article (which may or may not be the most important). Citations at least indicate whether others have found the work useful, one possible meaning of importance.

I write about what has come to be called the Deese–Roediger–McDermott (DRM) paradigm, owing to a suggestion by Endel Tulving. And Tulving is responsible in more ways than one. I had introduced him for a talk at Rice University in 1993 while I was on the faculty. The talk was on “the brain's proclivity for primacy” or why first (and new) events are well remembered. After my introduction, I sat off to the side. The talk was exciting, and the audience stayed long afterwards, asking questions. As moderator, I thought maybe they were going on too long, so I was about to pull the plug. Happily, I waited. Someone asked Tulving a question (the person and the question are lost to me), and during his reply Tulving said: “Your comment reminds me of an experiment by James Deese from years ago in which he presented lists of words, asked people to recall them in any order, and found they often intruded a related word in their recall” (my reconstruction of his answer, of course, not his exact words).

I wrote down “Deese experiment” on a note card. Later, I asked Tulving what year it might have been. He said “probably 1959.” He and I talked about how unusual that finding was, because free recall of words is often surprisingly error free.

The note got buried in a pile on my desk. A couple of months later I found it while cleaning up. This was before all the journals were online, so sometime later I trekked to the library and copied two Deese papers from 1959. One was the correct one (it had rarely been cited). I discovered that Deese had created lists of related words, words that were associates to a key, seed word. For example, he took the top associates to the word sleep (bed, rest, awake, tired, dream, etc.) for one of his lists.