Does the Gay Accent Exist?
Total Page:16
File Type:pdf, Size:1020Kb
Does the gay accent exist? An acoustic comparison between homosexual and heterosexual Dutch male millennials Quincy Liem 11050330 under the supervision of Dr. Silke Hamann A thesis submitted for the degree of BA in Linguistics June, 2019 TABLE OF CONTENTS ABSTRACT. .................................................................................................................................................... 2 INTRODUCTION ............................................................................................................................................. 2 DATA COLLECTION ........................................................................................................................................ 4 Participants ....................................................................................................................................................... 4 Procedure .......................................................................................................................................................... 5 DATA ANALYSIS ............................................................................................................................................. 5 Data cleanup ..................................................................................................................................................... 5 /s/ spectral peak frequency ............................................................................................................................... 6 Pitch and intonation .......................................................................................................................................... 7 Speech tempo .................................................................................................................................................... 8 Statistical analysis ............................................................................................................................................. 8 RESULTS ........................................................................................................................................................ 8 /s/ spectral peak frequency ............................................................................................................................... 9 Pitch and intonation .......................................................................................................................................... 9 Speech tempo .................................................................................................................................................. 10 CONCLUSION ................................................................................................................................................ 10 REFERENCES ................................................................................................................................................. 12 APPENDIX A. Abstract in Dutch .................................................................................................................... 15 ABSTRACT. ................................................................................................................................................... 15 APPENDIX B. Participants ............................................................................................................................. 16 APPENDIX C. Noise filter script ..................................................................................................................... 17 APPENDIX D. SKewness and center of gravity script ...................................................................................... 21 APPENDIX E. Pitch script ............................................................................................................................... 23 APPENDIX F. Speech tempo script ................................................................................................................ 24 APPENDIX G. Statistical analyses .................................................................................................................. 28 1 Does the gay accent exist? An acoustic comparison between homosexual and heterosexual Dutch male millennials Liem, Q.C., supervised by Hamann, S. ABSTRACT. In response to a perception-based approach employed in previous studies for English speaking gay men, a production-based experiment was conducted, focusing on Dutch gay men. Two groups were compared: Dutch heterosexual and Dutch homosexual men. Natural speech data elicited through informal sociolinguistic interviews were acoustically analyzed in Praat. No evidence was found to conclude that Dutch gay men's speech is distinctive from heterosexual men's speech based on the acoustic features which were compared in this study: skewness of the spectrum and center of gravity of the spectral peak frequency of the fricative /s/, mean pitch, pitch range, intonation (pitch variability over time), speech rate, and articulation rate.1 INTRODUCTION which has been reiterated by many researchers, is that gay men sound effeminate or that gay The 'gay accent' – does it really exist? Even speech is characterized by being imitative of those without a self-proclaimed gaydar often female speech (Travis, 1981; Gaudio, 1994; claim to be able to judge whether someone is Munson, DeBoe & White, 2006). But what does gay just by hearing the speaker's voice. The this mean? What are the key distinctive widespread cultural stereotype depicts an features for gay and/or feminine speech? effeminate voice which includes a high pitch, lisping, over the top intonation, and drawn out For example, women's voices have a pitch on vowels as salient markers of homosexuality in average 100-150 Hz higher than male voices men. But is this accent indeed just a stereotype (Smyth, Jacobs & Rogers, 2003; Morrison & or is there actually something tangible about Rogers, 2000), and Lakoff (1990:204) mentions the way that gay men speak that sets them women's more varied intonational contours. But apart from their straight peers? contrary to the stereotype, Munson & Zimmerman (2006) and Munson & Babel (2007) A sizeable body of research has gone into did not find evidence that gay male speech identifying and describing the key mimics female speech. They found no overall characteristics of this particular manner of differences in mean pitch between heterosexual speaking in English. The prevailing stereotype, and homosexual men. More extensive 1 The Dutch version of this Abstract can be found in Appendix A. 2 Heterosexual Homosexual Fig. 1. Spectrograms of the fricative /s/. The spectrogram of a self-identified homosexual man (right) has a more negative skewness (skewness = −1.67) than the spectrogram of a self- identified heterosexual man (skewness = −0.03). (Munson, DeBoe & White, 2006: 215) investigations regarding pitch and sexual more negative skew in the sibilant spectrum orientation were done by Gaudio (1994), and (Fig. 1). Linville et al. (1989), Rogers & Smyth Smyth, Jacobs & Rogers (2003). They found no (2003), Munson, Jefferson & McDonald (2006), evidence for a perceptual correlation between and Munson, DeBoe & White (2006) sexual orientation and pitch range or pitch corroborate the finding that an /s/ produced variability. In contrast, Levon's (2007) with a higher peak frequency and thus a more empirical research showed that listeners were negatively skewed /s/ is perceived as gayer- more likely to judge recorded speech stimuli, sounding. This distinctive /s/ appears to be a which were digitally manipulated to feature a hyper-articulated /s/, according to Maniwa, wider pitch range, as sounding gay. However, he Jongman & Wade (2009) and Munson & concedes that the findings are highly contingent Zimmerman (2006). and dependent on sociolinguistic factors. Similar results have been found in Danish. In a Panfili (2011) examined the relationship perception-based matched guise experiment between vowel duration and perceived sexual conducted by Pharao, Maegaard, Møller & orientation by letting listeners judge a pre- Kristiansen (2014), listeners were asked to rate recorded passage. This, however, did not yield Danish speech samples from two different social a solid correlation. registers into which either an alveolar /s/ or a fronted /s/ were spliced on scales of To investigate the stereotype of the 'gay lisp', homosexuality and femininity. The results Mack & Munson (2010) analyzed listeners' showed that for the modern Copenhagen speech perceptive associations of phonetic variations of register, listeners rated the speech samples the fricative /s/ in English with perceived containing a fronted /s/ higher on both the sexual orientation. They found that a sharper homosexuality and femininity scales compared /s/ that was articulated more frontally in the to the alveolar /s/. Interestingly, the fronted mouth was perceived as 'gayer-sounding'. This and alveolar /s/ did not receive any particular realization of the /s/ is realized significantly different ratings when they acoustically as a sibilant with an especially high occurred in the street language register. As the spectral peak frequency, which corresponds to a 3 fronted /s/ has been found to be indexical of In short, by comparing a random group of homosexuality in both English and in Danish Dutch homosexual men to a random group of based on perception, could we expect similar Dutch heterosexual men, can we objectively find findings in