Musicological Editions and the Textual Fallacy: On the Critical Use Of Musical Texts

Foucault observes that one cannot theorize about "editions" without first having a theory of "works." However, the edition-work relationship operates in both directions: how one encounters works in the editions one uses influences one's understanding of those works. Ever since the 1850s, when German publishers issued the first musicological editions- the first large-scale historical editions prepared with text-critical meth- odology-those editions have been the most respected form in which musical compositions are encountered. But those editions were products of a particular moment in the history of Western music, a moment when musical works were stable entities defined in great detail by composers who recorded their conceptions in texts that performers are expected to follow faithfully. With such a model, a composition might reasonably be represented by a single text, a fortunate circumstance, since offering more than one text of a work is economically impractical. That model, however, is not applicable to much music composed before the mid -eighteenth century, when texts were used differently and compositions were not necessarily defined by composers' texts. Musicological editions therefore represent much earlier music in ahistorical and therefore mis-leading ways.Recent textual theory suggests that, especially for earlier Western music, more than one text may be needed to represent a composition adequately. Digital editions may one day help address this by offering multiple texts of a composition, but meanwhile users of print musicolog-ical editions must approach them recognizing that they may obscure the different ways in which musical texts were used in the past.