During the first half of the 20th century, most theories of behavior i
nvoked some concept of motivation that included an energy or drive var
iable. For a variety of reasons, such variables were widely criticized
and have been generally out of fashion during the past 30 years. This
paper examines some of the criticisms of energy variables, especially
as postulated by Lorenz, and suggests ways of overcoming many of the
objections raised. An energy variable is proposed that has multiple so
urces of energy (both internal and external) and multiple means of dis
sipating energy (performance of the behavior, performance of alternati
ve behavior, and passage of time). Results from studies on dustbathing
in chickens, attack and other behavior in fish, and sleep in humans a
re reviewed and shown to be partially accounted for by fluctuations in
an energy variable. In most cases, however, an additional variable, c
alled threshold, is required for a full account of the observed behavi
or. It is concluded that a general theory of motivation should include
both an energy variable and a threshold variable. (C) 1997 Elsevier S
cience B.V.