Wisconsin lies at the edge of many of the most significant changes currently underway in American English, including the spread of a change in the vowels in cot and caught in which they come to be pronounced the same. (Article) This has reached western Wisconsin and is widespread to our south. Another larger set of vowel changes called the ‘Northern Cities Shift' (where, among other changes, Ann is pronounced increasingly as in p ian o ) has spread from New York as far as Madison. (Map)
Wisconsin also hosts older boundaries such as pop vs. soda, separating eastern Wisconsin from areas to the west. These and many other patterns reveal that Wisconsin English is remarkably distinctive, but the reasons for this distinctiveness are poorly understood: Until now, with few exceptions (most notably the late Frederic Cassidy's Wisconsin English Language Survey and Dictionary of American Regional English), the linguistic distinctiveness of our region has not been explored.
Samples of Vowel Patterns in Wisconsin
Current sociolinguistic work on American English has focused heavily on vowels and as a result, we know a fair bit about Wisconsin vowels. Let's look at this pattern in a little more detail.
As just noted, two major patterns of vowel changes meet in Wisconsin. The eastward change is where the words caught and cot are pronounced essentially the same. The westward change is where vowels rotate in what is called the Northern Cities Shift ( bit > bet > butt > bought > baht > bat; six > sex > sucks > Saux's > socks > sax ).
In order to demonstrate the variability within the state, one can take recordings of six subjects who were recorded by John Westbury in the early 1990s at the University of Wisconsin 's X-ray Microbeam Laboratory in Madison. These six subjects made two north-south dialect continua, one on the east side of the state from Green Bay (male, 19) to Kiel (female, 20) to Milwaukee (female, 20), and the other on the west side of the state from Amery (female, 20) to Blair (female, 20) to Madison (female, 19). Moreover, these subjects are interesting because the continua span the two dialect regions: the Inland North region (which includes Minnesota), and the Northern Cities region which spans the crescent below and between the Great Lakes from Madison, WI, to Syracuse and Rochester, NY (excluding some cities like Erie, PA, in between). The map below shows the locations of the subjects.
Figure 1. Map of Wisconsin vowel dialect regions. The white region represents the Inland North region, the darker red represents the Midland region, and the lighter red represents the Northern Cities region. The white star represents Madison , and the star plus the five blue diamonds represent speakers' dialect locales.
The subjects were recorded saying the main vowels of American English sandwiched between a /s/ and /d/ (e.g., said, sid, sayed ). Vowels are categorized by linguists in terms of where the main constriction occurs in the mouth, so that the vowel in seed is considered a “high, front” vowel, while the vowel in sod is considered a “low, back” vowel. Approximating these relative positions on a graph is fairly easy. One simply needs to have a tool such as a software package to calculate the resonances of the digitized sound file. The figure below is a graphic representation of some of the vowels of the six speakers using a psychoacoustic scale (the Bark scale (Z)) which reflects both the linear and logarithmic aspects of the auditory system. Moreover, we can identify the vowels by comparing the lowest three main resonances in the mouth (called formants, and labeled F1 for the lowest formant, F2 for the next one, and so on) and subtracting the first and second from the third. The figure below shows the subjects by dialect region (Inland North v. Northern Cities), and the figure immediately below that shows the subjects by continua (eastern and western).
Figure 2. A graphic representation of the vowels of the six speakers by dialect region. The top panel represents the three speakers from the Inland North region, with the red vowels identifying the speaker closest to Minnesota. The bottom panel represents the three speakers from the Northern Cities region, with the red vowels representing the speaker from Milwaukee in the southeast part of Wisconsin.
Figure 2 is interesting because it shows that for the Amery subject, the caught~cot merger is complete: the two vowels are sitting in the same spot. Likewise, this subject has a fairly low vowel in sad , and especially lower and forward than the vowel in said. In contrast, the subject from Milwaukee has a distinct separation between the vowels in all four words. Yet, the vowels in sad and sid are quite close, with sad well above said (which looks like it is further back in the mouth than sod ).
Figure 3. A graphic representation of the vowels of the six speakers by geographic continua. The top panel represents the three speakers on the western side of the state, with the red vowels identifying the speaker from Madison. The bottom panel represents the three speakers along the eastern, with the red vowels representing the speaker from Green Bay.
Figure 3 is also interesting. In the top panel we see that the Madison speaker, with respect to the Inland North speakers has the vowel in said in the midst of the others' sad , and the vowel in sod is more forward than the others. In the bottom panel, the Green Bay speaker's vowels—in the absence of other evidence—argue for raising the dialect boundary a little bit higher. This claim can be made in light of the raised vowel in sad. Note as well that the caught~cot merger is not represented in this speaker's vowels.