In this study, two groups of students were asked to listen to recordings made of Senegalese Wolof speakers and make deductions about their social and caste status. The responses of the first group, made up of Senegalese students, did not go beyond the threshold of chance with regard to caste status, but were 65.7% correct regarding the speakers' social status. The second group, who were European students with no prior knowledge of the Wolof language, achieved percentages of correct answers similar to those of the Senegalese listeners with regard to social status. The imposed norm hypothesis, which predicts that sociolinguistic features cannot be gauged by those who have had no previous contact with the community, should thus be reconsidered and enlarge its scope to include a more general, and therefore nuanced, view of language. (Imposed norm hypothesis, inherent value hypothesis, social stratification of language, social identification, Wolof, Senegal, castes)*