William Labov and The Social Stratification of R
William Labov and The Social Stratification of R
Hypothesis
In his 1966 study, William Labov hypothesized that the pronunciation of the /r/ sound
was a social differentiator across the population of New York City. The general hypothesis was
that if two groups are ranked by the systematic differences in their social status or prestige, then
they will be ranked similarly by their differing use of /r/. Labov tested this theory by
interviewing salespeople at three different New York City department stores, each associated
with a difference social status. He predicted that salespeople in the highest-ranked store, Saks
Fifth Avenue, would have the highest value of /r/, those in the middle-class store, Macy’s, would
have intermediate values of /r/, and those in the lowest-class store, S. Klein, would have the
lowest values.
It was assumed that salespeople at these three stores had the social characteristics of the
respective classes that the stores catered to; salesgirls at the highest-ranked store took on the
prestige of their customers. The stores were clearly stratified by number of advertising pages in
different newspapers, prices and how prices are displayed, and spaciousness and layout of the
store itself. Labov also hypothesized that observation of speech in a natural context, rather than a
Method
The interviewer, Labov himself, observed employees of each store saying the words “fourth
floor” by posing as a customer asking for directions to a section he knew to be on the fourth
floor. Upon asking for clarification, the interviewer would receive another utterance with more
careful style and emphatic stress. Labov then noted the independent variables of which store, the
floor within the store, the age (within a five- year estimate) and sex of the salesperson, their
specific occupation within the store, their race, and any foreign or regional accent.
The dependent variable was the use of /r/, and the interviewer obtained four occurrences of the
sound per salesperson: “fourth floor” said casually, and “fourth floor” repeated emphatically.
Also noted were any other occurrences of /r/ in the speech of the informant throughout the course
of the interview. For full pronunciation of /r/, (r-1) was marked, and for each non-constricted
This was repeated on every aisle of every floor in the entire store as many times as
possible without the informants noticing that the same question had already been asked. On the
fourth floor, the question asked was “Excuse me, what floor is this?” By utilizing this method of
interviewing via a casual, normal, salesman-customer exchange, the informants spoke with their
usual style of speech. Results were not affected by the perceived pressure of a formal interview,
and the occurrences of /r/ were observed in the desired social context.
Overall Stratification
As hypothesized, the salespeople were ranked in the same order based on their use of /r/ as
the social class of their respective stores. The percentages of store employees who pronounced /r/
fully without exception (r-1) rose from 4% at Kleins, to 20% at Macy’s, and 30% at Saks (Figure
13.1). Conversely, when comparing data for complete responses only, the percentages of those
who did not pronounce /r/ (no r-1) fell from 82% at Kleins, to 41% at Macy’s, and 30% at Saks
(Table 13.2). The data shows clearly that the store with higher social prestige had higher rates of
employees pronouncing /r/ fully, and lower rates of /r/ being omitted, while the opposite was true
employees at both Macy’s and Kleins. The majority of Macy’s employees aimed for r-
pronunciation, and instances of r-1 were higher in their emphatic speech, but it was not the
method of speech that they used as a norm. Rates of Macy’s employees pronouncing /r/ rose to
nearly the same level as those from Saks, which was the highest increase among all three stores.
For Kleins employees, the percentage of salespeople pronouncing /r/ rose from 5% to 18% as the
context became more emphatic. Saks employees, however, showed a much less marked
difference in r-pronunciation as speech became more emphatic, which suggests that those in a