Recent studies, by us and others, have argued that the Second Consonant Shift began medially after stressed short vowels, triggered by a segmental interpretation of aspiration in interaction with Germanic syllable weight requirements. The most striking empirical support came from the dialect of Wermelskirchen, where shift of fortis stops is attested only following short vowels. But is Wermelskirchen an isolated dialect or part of a general pattern? We review selected dialect data supporting this new account of the shift and show the Wermelskirchen evidence to be cut from a broader regional fabric that is marked also by biases in place of articulation among stops and, to some extent, their following vowels. We take these data to reflect the archaic nature of the modern distributions, concluding that the apparent idiosyncrasies obscure an original, fundamental regularity whose structural motivations come into clearer focus under the principles of Evolutionary Phonology.An earlier version of this paper was presented at the 11th Germanic Linguistics Annual Conference (GLAC-11), University of California–Davis, in April 2005. In addition to members of that audience and two readers for this journal, we have had the privilege of valuable discussions on this topic and comments on earlier drafts from the following colleagues, none of whom necessarily agrees with all of our points: Juliette Blevins, Markus Denkler, David Fertig, Patrick Honeybone, Robert B. Howell, Thomas Klein, Mark Louden, Jürgen Macha, Monica Macaulay, Klaus J. Mattheier, Donka Minkova, Richard Page, and Laura Smith. We also thank Garry Davis for sharing a copy of the handout from his 2003 presentation with us.