This paper pursues an “ingenerate” or phonetically based account of i-umlaut as it unfolded in North Germanic. We focus on a famous problem relating to umlaut distributions in i-stem nouns: In the long stems of that class (gestr ‘guest’, from earlier *gastiz), where umlaut is arguably less motivated phonetically, it is generally reflected throughout the paradigms, but in short stems (staðr ‘place’, from earlier *staðir), where it is more expected, umlaut is generally absent. A central feature of our understanding of these and other Norse facts is the interleaving of processes of sound change and analogy, the latter of which, by an assumption validated elsewhere, comes into play only under extraordinary circumstances. In contrast to previous work on the conundrum of umlaut in Old Norse, we situate this account in the context of umlaut as a general phenomenon, with parallels in development to that of its West Germanic sisters.Besides many members of the audience at the XVIth International Conference on Historical Linguistics (Copenhagen, August 2003), where a preliminary version of this paper was presented, we thank two anonymous readers for this journal and the following colleagues for their comments and discussion: Anthony Buccini, Rob Howell, Monica Macaulay, Richard Page, Michael Schulte, J. C. Smith, as well as Laura Catharine Smith and other members of the UW Phonology Group. As may become clear in the course of the paper, these individuals do not necessarily agree with what follows; of course, any shortcomings remain our own. The translations of Danish and Norwegian quotes are likewise our own.