290 Thomas C. Purnell
Ethnicity is generally a moving target for researchers because, among other
reasons, ethnicity often transcends geo-political boundaries. Examples
related to the ethnic groups examined in this study include American Eng-
l ish sp e a ke r s w it h t ie s t o p e rceive d A f r ica n or G e r m a n a nce st r y fou nd a c ross
many geographic locations. Another indication that ethnicity is a potential
illusion is that ethnic group membership has À uid temporal and situational
characteristics. Baugh (e.g., 1988) found that consonant cluster reduction
among African American Vernacular speakers is inÀ uenced more by famil-
iarity than ethnicity. Nonetheless, ethnically af¿ liated speech is afforded
some perceptual substance since ethnicity can sometimes be attributed to
geographic regions, e.g., German in Lancaster County, Pennsylvania, and
southeastern Wisconsin. Additionally, characteristics of ethnic speech pat-
terns attain a degree of perceptual reality as seen in metathesis and cluster
reduction in African American English (Rickford 1999 among many oth-
ers). Finally, the “reality” of ethnically af¿ liated dialects in the US con-
tributes to social problems dealing with race and ethnicity (e.g., “linguistic
pro¿ ling,” Baugh 2000), and reÀ ects shared knowledge between speakers
and listeners.
It makes sense, then, to place the perception of ethnically af¿ liated dia-
lects within the context of a paradigm accentuating the role that the listener
has in the preservation of speech patterns. Perceptual dialectology (Preston
1989, 1997, 2002 et al.) allows for such a focus by de¿ ning dialect as listener-
centric and social boundary-sensitive. It is listener-centric because it is “an
assessment of... ‘attitudes’” (Preston 2002: 95). This additional notion of
dialect as an abstract, perceptual category adds to many traditional notions
of dialect.^1 Perceptual dialectology is also sensitive to folk speech boundar-
ies and is in line with the concept of dialect as a social boundary index. At
the core of perceptual dialectology is the observation that research on how
people react to others’ speech or what they say about a speaker because of the
speaker’s dialect is just as important as what is uttered (Preston 1997, 2000).^2
This orientation to linguistic research is illustrated in Figure 13.1.
Preston’s intention in his presentation of Figure 13.1 is to depict not only
the speech act (a) and reactions to that act (b and c in the inner triangle),
but also linguists’ prioritization of research based on what is going on when
someone talks (the outer triangle). Preston (and Hoenigswald before him)
acknowledges that linguists have been preoccupied with the act of speaking
(a’) and with determining who and who is not an authentic or legitimate rep-
resentative of some such speech community based on their speech. Instead,
Preston argues that listeners cluster speakers into folk or cognitive catego-
ries by what is said (c’ and c) and how they react (b’ and b), and that this