Sociolinguists used to think that dialect differences were primarily based on geography. They would ask people to say words from a list and try to identify isoglosses (sharp geographical boundary lines) based on differences in pronunciation. However, there were problems with this methodology. First, linguists often found that a clear isogloss didn’t exist. Instead, the boundaries were fuzzy and contained considerable overlap. Second, much of the population is very mobile, resulting in heterogeneous dialect in many regions. Schmidt (1872) claimed that linguistic changes spread across the country like waves. "Different innovations spread at different rates and perhaps in different directions, so that the areas covered by different innovations do not coincide. The result of successive waves is a network of isoglosses" (Wells 1982; 13). This idea was known as the wave-theory. Trudgill (1974) challenged the wave-theory, claiming that instead, linguistic innovations spread from cities to larger towns, then from larger towns to smaller towns, and finally, from smaller towns to the countryside.

Then a sociolinguist named William Labov came along and radically changed notions regarding dialect differences. He illustrated that linguistic variation is not only geographically stratified, but it is socially stratified as well. One of Labov’s most famous studies is of New York department stores. He analyzed the rhocity of employees who worked at three department stores in New York—Sack’s, Macy’s, and Klein’s. His methodology consisted of asking employees on each floor of each department store where ladies coats were located in order to elicit the response "fourth floor." His findings were interesting. First, he discovered that r-fulness increased as he went up the floors of the department stores. So the first floor employees had the least rhocity and the fourth floor employees had the most rhocity. This corresponds with the price of items on each floor, with the least expensive items being located on the first floor and the most expensive items being located on the fourth floor.

Second, Labov found that rhocity increased with the prestige of the department store. So, if we ranked the department stores based on prestige, Sack’s would be first, Macy’s second, and Klein’s third. Accordingly, Sack’s employees had the most rhocity and Klein’s employees had the least rhocity. These findings were important in identifying class and socioeconomic status as an index of dialect.

Labov also distinguished between careful and casual speech in this study. He found that people’s first responses were often more casual than their second responses. Labov discovered this using the following methodology: He asked employees where ladies coats were located. When they replied, "fourth floor," he would ask, "what did you say?", thereby eliciting another response of "fourth floor." He discovered that the first response (casual speech) was often r-less, while the second response (careful speech) was then r-full.

In this unit, we will focus on perceptions and attitudes regarding sociolinguistic variation. The excerpt that you will read from Eckert’s Linguistic Variation as Social Practice (2000) provides a contextualized background for the study of sociolinguistic variation and insight into current variation study trends. Purnell, et. al.’s article illustrates descrimination that results from social/racial stratifications of dialect. The authors analyze housing discrimination based on dialect, a feature of speech that indexes ethnicity. They claim that callers in their study made determinations about a person’s ethnicity as soon as the word "hello" was uttered into a telephone.

READINGS:

• Eckert, Penelope, "Interpreting the Meaning of Variation," Linguistic Variation as Social Practice, chapter 1, pp. 7-45.
  

• Purnell, Thomas, Idsardi, & Baugh, "Perceptual and Phonetic Experiments on American English Dialect Identification," Journal of Language and Social Psychology, 18.1, pp. 10-30

DISCUSSION:

There are many levels of variation in language. Lexical variation is at the level of the word. Examples of this include "gumband" vs. "rubberband," "curb" vs. "burm," and "median" vs. "esplinade." Syntactic variation is at the level of the sentence and involves the actual structure of how the language is put together grammatically. An example of this is "going to" vs. "fixing to." Pragmatic variation is at the level of social practice and involves the unwritten rules of social interaction. An example of this is "Can I have a drink of your soda?" vs. "Let me get a drink of your soda." Pragmatic variation also includes rituals of offering food/drink to guests, saying goodbye, and teasing.

Can you think of other examples from your own personal experience that demonstrate each of the levels of variation discussed above (lexical, syntactic, pragmatic)? Which level do you think plays the greatest role in language attitudes and stereotypes?

What are some common language stereotypes that you are aware of? For example, "New Yorkers are rude" or "Southerners are slow." What levels of linguistic variation contribute to these stereotypes?

Think about what contributes to language attitudes in an even broader social sense? In other words, how is language political? Why do people feel invested in maintaining language stereotypes? What is in it for us?