Journal of Sociolinguistics 13/5, 2009: 613–633 Martha Stewart behaving Badly: Parody and the symbolic meaning of style 1 Jennifer Sclafani Hellenic American University, Greece This study addresses the issue of how to correlate social meaning with linguistic style through an investigation of the parodic speech genre. The analysis examines two parodies of lifestyle entrepreneur Martha Stewart and compares linguistic strategies used in parodies of Stewart to her own linguistic performance on her talk show. Features considered include phonological characteristics, lexical items, politeness strategies, and voice quality. A comparative quantitative analysis of aspirated and released /t/ as employed by Stewart and her parodist reveals that a variable feature of Stewart’s style is rendered categorical in the parody. It is demonstrated that both parodies exploit elements associated with Stewart’s ‘Good Woman’ image in order to expose Stewart as a ‘Bad Woman’, a reputation she earned for her 2003 insider trading conviction. This study suggests that parodic performance may serve to strengthen and even iconize indexical connections between stylistic variants and their social meaning in particular contexts. KEYWORDS: Martha Stewart, style, parody, indexicality, media discourse, sociolinguistics 1. INTRODUCTION The notion of stylistic variation has evolved throughout recent decades from being seen as a primarily reactive phenomenon, as exemplified in Labov’s (1972) ‘attention to speech’ model, to being viewed as a more proactive display of identity in interaction (Bell 1984, 2001; Coupland 2001a, 2001b; Eckert and McConnell-Ginet 1995; Schilling-Estes 1998). Related to this paradigmatic shift toward viewing style as a more agentive rather than reflective phenomenon is the question of whether a speaker’s stylistic shifts can be best understood as a response to the speaker’s audience or as an initiative and strategic move. Bell (1984, 2001), who draws on Bakhtin’s (1981) dialogic theory of language, emphasizes that the responsive and initiative dimensions of style-shifting should be viewed as two sides of the same coin, and that ‘responsiveness to the audience is an active role of speakers’ (Bell 2001: 143; emphasis mine). Sociolinguistic research has provided much evidence for this view of style-shifting, often employing the method of observing individual speakers in multiple interactional C  Blackwell Publishing Ltd. 2009 9600 Garsington Road, Oxford OX4 2DQ, UK, and 350 Main Street, Malden MA 02148, USA