In the spring of 1934, after I had first played the Goldberg Variations in New York, Arthur Mendel suggested editing them and brought me together with Carl Engel, editor for the publishing house of G. Schirmer in New York. During a subsequent transatlantic voyage I began preparing the text, and that summer in Salzburg wrote most of the preface. I was then twenty-three years old, and full of zeal for purity of texts that had little precedent in the so-called practical editions hitherto published in America, with the notable exception of Schirmer's own Widor-Schweitzer edition of Bach's organ works. My views were not always those held by Mr. Engel, or by Harold Bauer, the eminent pianist who often served as consultant to Schirmer's. But by agreeing to relatively harmless interventions that today would probably not be considered necessary, such as indicating the execution of ornaments and writing out crossing voices on extra staves, I managed to leave the text free of any such accretions as fingerings, phrasings, dynamics, or tempo markings. The volume was finally published in May 1938 with a handsome cover that reproduced the engraved border from the title page of Bach's original edition.

In 1956, I was asked whether I wished to make any alterations in my preface for the printing of an English edition, and I replied that it might be best not to intervene. I already regarded this preface as a document of my extreme youth that might be hard to match with the ideas and style of later years, but I did not yet foresee that I would one day find still more with which to disagree and which I would wish to amend.

I first played the Variations for a music class when a senior at Harvard in 1931, then for the first time in public in Berlin in January 1933. My file of surviving programs indicates at least ninety-nine performances through June 1970. I had twice recorded them, once in 1952 and again in 1958. All these performances were without repeats. The performances varied and the interpretation changed, but over the years certain notions remained constant—for example, the desire to present the work as a complete whole and to respect the rhythm and proportions with which its component parts are arranged.