Glenn Gould's recording career is bookended by his 1955 and 1981 recordings of Bach's Goldberg Variations. Gould discussed these two performances at some length during interviews around the time of the 1981 release, and in these comments he expounded a loose theory of a "constant rhythmic reference point," the organizing principle behind the time dimension of his 1981 recording. Gould maintained that this aspect of the latter recording made it superior to his earlier effort by giving unity to

more »

... e set as a whole. Three excerpts from both recordings were included as part of an empirical study on tactus choice. To discover whether Gould was successful in communicating this unity to the average listener, these excerpts were taken from transitions between adjacent variations. While participants' tactus choices across these transitions were not uniform in response to either recording, they were much less diverse in response to the 1981 performance. Further, participants' tactus connections in response to the 1981 recording largely matched those that Gould explicitly sought to make. The results suggest individual and combined effects of Bach's composed metric structure and Gould's performance decisions relative to that structure, and indicate that Gould was (and is) able to control listeners' perception of musical time via their tactus to a greater extent with his 1981 Goldberg Variations.