258 Nancy Niedzielski
vowel merger prenasally (as illustrated previously in the Koops et al. study)
in Houston, Texas was recorded while subjects heard a sentence like “Sign
the check with a pen” or “Sign the check with a *pin.” For non-merged speak-
ers, the second sentence is ungrammatical, but it is acceptable for those that
do not make the pin/pen distinction. Those subjects with the merger exhib-
ited a reduced late positive event-related potential component, compared to
those without the merger, in response to hearing potentially ungrammatical
sentences. This suggests that implicit phonological knowledge affects even
neural activity (or perhaps vice versa), and suggests that such knowledge has
a demonstrated neurological component.
This set of research on implicit knowledge of speaker and language sug-
gests a mechanism by which sociolinguistic indicators—that is, variables
below the level of conscious awareness—can spread through a given speech
community with the speed and systematicity seen in variable after variable
(such as the regular 100 Hz raising of (aw) in Philadephia by decade, shown in
Labov [2001]). Shadowing tasks reveal that speakers accommodate towards
even low-level variation immediately, and that there is greater accommoda-
tion with increased numbers of repetitions. The research on mergers shows
that there is implicit knowledge of the correlation between, in these instances,
age and variation, even if subjects do not reveal this knowledge explicitly.
Thus, contact can produce immediate accommodation, which with length-
ened and stronger ties can become more permanent, and, furthermore, people
have intricate knowledge of how variation patterns in their community, even
if they are not able to consciously express this knowledge.
- Research on explicit knowledge
Researchers from various different ¿ elds have used several different meth-
ods in the investigation of language attitudes and language ideology. For
instance, the work of perceptual dialectologists (e.g., Preston 1989, Preston
1999, Long and Preston 2002, and the authors therein) on US English pro-
vides information about non-linguists’ impressions of dialect boundaries,
regional de¿ nitions of standard English, and the regions in which respondents
believe that “correct” English is spoken. Social psychologists (e.g., myriad
studies including and since Lambert 1967, Giles 1970, Giles and Powesland
1975) investigate attitudes towards speakers of various dialects, and the per-
ceived standardness of those dialects is revealed through the degrees and
types of prestige assigned to their speakers. Language attitude work (e.g.,
Lippi-Green 1997) has provided ample information about the forms that are