The tendency for information learnt in a particular mental or physical state to be most easily remembered in a similar state. The phenomenon was first investigated systematically in 1964 by the US psychologist Donald (Albert) Overton (born 1935), who focused on drug-dependent memory in rats, and this was later demonstrated in humans: material learnt in a drunken state is best remembered in a later drunken state, and so on. Clinical evidence has been cited of heavy drinkers who hide alcohol or money when drunk and are unable to find it when sober but remember where it is as soon as they become drunk again. It has since been established that many classes of drugs produce the effect, that the effect is dose-dependent, that it affects recall but not recognition, and that non-pharmacological states can also elicit the effect demonstrated: in a frequently cited British experiment on context-dependent memory, lists of words that were learned by divers while they were underwater were best recalled when they were again underwater, and conversely words learned on land were best recalled on land. Emotional states can also produce the effect—mood-dependent memory—and this helps to explain why pleasant experiences are more likely to be remembered by a person who is happy, and unpleasant experiences by someone who is unhappy and is likely to become even more unhappy as a result. Also called state-dependent learning. See also cognitive interview, encoding specificity, redintegration.