This work focuses on phonetic phenomena in fluent vernacular speech. For a definition of the vernacular, see Labov (1966). In short, it is an un-monitored speech style, a style in which the speaker pays attention to the content rather than to phonological or grammatical form. Thus the speaker is caught up in what he or she is saying, and pays little attention to how he or she is saying it. The differences between this and more formal styles of speech in some segments of a community are taken to an extreme in the speech of certain social groups, which are in this sense the most vernacular speakers. For example, in New York City, increased r-lessness is associated with informal styles for individual speakers, and also occurs at the highest rate among working-class speakers. Vernacular /r/ behavior in New York is both a pattern of style-shifting and a pattern of social variation (Labov 1966)
Why study vernacular, fluent speech? What are the differences between the vernacular and ``laboratory speech''? Laboratory studies frequently use fully-stressed citation forms, often single-word utterances making up a complete intonational phrase; they have often been rehearsed by the speaker well in advance of the production; they may be nonsense words; they are typically read from a prepared text or memorized. Yet these factors may obliterate regular phonetic conditioning, as shown by Keating and Huffman (1984) and Labov (1986). The effect of self-reflection on one's speech is to introduce unsystematic ``correction'' of the phonetic forms produced -- which amounts to bias in the data. On the other hand, the regularity of phonological patterning in vernacular speech as opposed to formal, or highly self-monitored speech, has been persuasively argued for (cf. Labov 1966, 1986). Further, fluent vernacular speech is language's most basic, systematic, as well as frequent form. The vernacular is fundamental to the history of languages, since it is the basis of their historical continuity. The vernacular is fundamental to the social meaning of language forms, since speakers are (socially) defined, generally for life, by the vernacular dialect acquired in their youth.
Eckert (1989) and others show that sound change takes place in the young (adolescent) peer group. In the second decade of life, the phonetic forms that speakers will continue to produce regularly throughout their lives are molded through the heat of peer pressure (or perhaps by the burning desire to conform), a process which also gives sounds their most deeply-felt social meanings. It is the local vernacular dialect which is acquired in this period, and which is itself gradually modified by both social and linguistic forces of change.
Thus, the vernacular local dialect is the locus of sound change. It might be argued that only a dialect that has historical continuity is ``real,'' or historically significant. Consider a somewhat artificial example: suppose (incorrectly) that (what is normally thought of as) Latin and French are directly related as ancestor to descendant, i.e., that they were historical stages of a single, continuous, living language. The synchronic and historical studies of phonological variation and change by Labov and his students and colleagues may be extended to establish the Latin/French vernacular as the site of the changes that differentiate these historical stages. In this case, the true French and the true Latin, which are related concretely, historically, are vernacular French and vernacular (spoken, ``Vulgar'') Latin. Book-Latin and book-French are not related in this direct, concrete way, and are therefore mere historical sideshows, dead remnants of the historically continuous, vernacular, living language.
Thus it is the vernacular that is the historically significant, or ``real'' form of language. This a priori historical argument has added merit if we consider the associated facts that most people are vernacular speakers, that local prestige of individuals is correlated with their degree of advancement in use of the changing vernacular (so that the local people that are most important, in some sense, are the ones most advanced in their speech)4.2 and finally, that the most regular and systematic form of language is the vernacular (Labov 1966). For these reasons vernacular speech is an important object of linguistic study. The core of this work consists of phonetic results derived from the frequent, natural, systematically conditioned, socially meaningful sounds found in everyday, fluent, vernacular speech.