The playback experiment -- the playing back of recorded animal sounds to the animals in order to observe their responses -- has twice become central to celebrated researches on non-human primates. First, in the years around 1890, Richard Garner, an amateur scientist and evolutionary enthusiast, used the new wax cylinder phonograph to record and reproduce monkey utterances with the aim of translating them. Second, in the years around 1980, the ethologists Peter Marler, Robert Seyfarth, and Dorothy Cheney used tape recorders (...) in a broadly similar way to test whether the different predator calls of one monkey species, vervet monkeys, warn about different kinds of predator. This paper explores the circumstances leading to the ca. 1890 invention and the ca. 1980 reinvention of the primate playback experiment. In both instances, I show, the experiment served as a riposte to those arguing, on scientific grounds, that an unbridgeable gap divides human language from animal communication. I also consider how far progress in technology explains the timing of invention and reinvention. I conclude with some reflections on sifting contingent from inevitable aspects of the history of the primate playback experiment, and of scientific achievements more generally. (shrink)