Tales of the Northern Cities

In recent years, the study of American dialects has increasingly drawn the interest of sociolinguists, especially those working in the Labovian paradigm. As one of these sociolinguists, I hope it is not too presumptuous to suggest that this trend has rejuvenated American dialectology by attracting new scholars to the field and, more importantly, by expanding the range of issues studied and the methods used to explore them. One significant consequence of this development has been a shift in focus from studying the preservation of older linguistic features (e.g., regional vocabulary items like mosquito hawk and darning needle as words for 'dragonfly') to studying the spread of innovative features (e.g., sound changes like the "cot/caught merger"). This emphasis on "language change in progress" has opened new avenues of research that have increased our understanding of the structure and use of American English as well as of language more generally. By discussing my own experience studying a well-known change in progress, I hope to illustrate some of the opportunities to be found along these new avenues of research.

Much of my research has investigated a series of sound changes known as the Northern Cities Shift (NCS). The NCS involves the movement of six vowels and is heard across the Northern dialect region from western New England to the Dakotas (for details see Gordon 2000, forthcoming). The NCS has been made familiar to most students of American English through the work of William Labov. The first detailed discussion of the NCS appeared in Labov, Yaeger, and Steiner (1972), and Labov's research has continued to examine these changes (see, e.g., Labov 1994). When I first became interested in studying the NCS, just five years ago, I had some doubts about its viability as a topic of long-term research. Given that the NCS had been studied for over 25 years by eminent scholars like Labov, I assumed that the important work had already been done and that I would have little of interest to contribute. As I examined the literature on the NCS, however, I came to realize how mistaken my assumptions had been. While Labov and others had conducted excellent research and had examined many key issues pertaining to the NCS, there were significant gaps. These gaps included fairly straightforward questions, such as which phonological environments tend to promote the changes, as well as more challenging questions, such as whether the changes are causally linked in a chain shift (see Gordon forthcoming).

That such questions remained open after decades of research is a testament to the complexity of the NCS. After all, the NCS involves changes to six vowels and affects millions of speakers across a broad geographic region; it raises issues of interest to phonologists, sociolinguists, and historical linguists, as well as dialectologists. I suspect that the NCS is not altogether exceptional in this regard, that there are many other such complex phenomena active in American English today. The general lesson I hope researchers, especially students, will draw here is that we should not assume that a long line of research into a subject by leading scholars means that all significant issues related to that subject have been resolved. A more particular lesson, again especially for students, is that there remain many unresolved questions related to the NCS and many opportunities for contributions.

As an illustration, let me briefly consider the two questions that I am asked most often about the NCS, and to which I wish I had better answers: What started these changes and what is going to happen in the future? With regard to the former question, the internal mechanics of the NCS are complicated, and there is debate about which of the vowels shifted first. More surprisingly, even the geographic picture is unclear; we do not know where the NCS began. This origin question is a prime example of an area where the union of dialectology and sociolinguistics can be...