Language in Society


“We act like girls and we don't act like men”: Ethnicity and local language change in a Philadelphia high school

Suzanne Evans Wagner

Department of Linguistics and Languages A-632 Wells Hall, Michigan State University East Lansing, MI 48824, USA


How is ethnicity indexed linguistically in a speech community in which immigrant L2s have typically not been spoken for three or more generations? Drawing on recordings and ethnographic observations of eighteen white high school girls in south Philadelphia, speakers of Irish descent are shown to differentiate themselves from speakers of Italian descent through their use of (ay0), that is, Canadian Raising. (ay0) is an ongoing sound change in Philadelphia and is remarkable for being a rare example of a male-led change. Irish girls exploit more male-like, backed, and raised variants as a resource for indexing their ethnic identity, which is associated locally with stereotypically masculine characteristics such as toughness. The symbolic reflection of ethnic affiliation through this subtle linguistic device makes use of both local and supralocal social meanings. (Ethnicity, adolescence, Philadelphia, Irish, Canadian Raising, gender, sound change, language, and identity)*

(Received November 15 2011)

(Revised August 06 2012)

(Accepted August 28 2012)

(Reviewed September 13 2012)


*  This article derives from my doctoral dissertation work, the completion of which could not have occurred without the support of my committee: Gillian Sankoff, William Labov, and Penelope Eckert. I am especially grateful to Penny for encouraging this particular line of enquiry. The participants in the study gave generously of their time and opinions, and are more responsible than anyone for leading me into this exploration of modern American ethnic identity. I must also thank the school administrators who allowed me to spend time with their students. An earlier version of this article was presented at the 35th New Ways of Analyzing Variation conference in 2006 and at the 30th Penn Linguistics Colloquium in 2007. I am grateful to audience members at both events. Two anonymous reviewers provided very helpful comments. Any errors are my own.