Gender Sexuality and Bad Language Tony Mc Enery

Gender, Sexuality and Bad Language Tony Mc. Enery, Department of Linguistics and Modern English Language, University of Lancaster

The background to this talk • Work at Lancaster (Paul Baker, Andrew Hardie, Neil Millar) supported by a grant from the University • Lancaster Corpus of Abuse (LCA) • Published in part in ‘Swearing in English’ (2005) and a number of journal articles.

• Why - well there has been corpus based studies of swearing (Leech, Stenstrom, Ljung) • The main studies of swearing remain noncorpus informed: – Slang (Partridge, 1960) – Anatomy of Swearing (Montagu, 1967, 1973) – Female Eunuch (Greer, 1970) – Language and Woman’s Place (Lakoff, 1975) – Swearing (Hughes, 1991, 1998)

• All of these studies make claims about this form of linguistic behaviour which is amenable, to lesser or greater degrees to corpus study • In this talk I will focus on a few claims made by Hughes by way of illustration and then move on to examine gender related work • But first …. .

How are we doing it? • Using categorisations used by others to develop an annotation of all of the ‘swear’ words in the BNC spoken corpus(LCA 1. 0) and later a broader set of words (LCA 2. 0) • Some studies have not considered various forms of swearing (e. g. swear words in a premodifying position) so I developed a categorization of these

• Used Sara to mine data from the BNC using three variables as our search parameters sex, age, social class. Work had to be redone as corpus was corrected. • Of the three variables, the last was and remains problematic • Each word is then encoded to indicate the age, sex and social class of the speaker amongst other things

• Further annotation is added to reveal

There is plenty to do! u Let’s quickly look at some claims made by Hughes, then move on to look at gay, queer, puff and fuck. First Hughes: u Claim one - the categories of swearing. Which words fall into which categories u Claim two - which words are used to insult which sex

Looking at infrequent words • Some of the word forms we are looking for have a relatively low frequency in he corpus • Words related to sexual orientation are such words • The reasons for this are interesting to consider • Though small, the data sets may give interesting suggestions which may be followed up by web as corpus studies, for example

Gays, queers and poofs • Data is sparse • But even on a small scale the data is interesting • Gay (24 examples) • Collocates: Is (10), He’s (9), You’re (2), Dad’s (1), Who’s (1) • A prosody of attribution in nearly all of the cases (21)

• Strong colligation with the “X is gay” pattern. • The X is male: • He’s (9), he (3), chap (1), dad’s (1), Mick (1), James (1), Male (1), Pat (1), Phil (1), sons (1) • Interestingly, no personal attributions of being gay.

• Queer (3 examples) • One abusive, but two have negative attributions! • Similar pattern of colligation, but negation included • “X is not queer” • Is it that we are abusive of that we claim we are not?

• Poofter (6 examples) & poof (2 examples) • Singular common nouns. Always P abuse. • Not used in an attributive manner

• But note here that we are within a heterosexual (or at least nominally heterosexual) discourse community. This pattern could clearly change if we shift to a homosexual discourse community. • The data is insufficient to test the hypotheses, but it is a useful spur to the flank of the analyst, and can set a research agenda to be pursued by other means

Looking at more frequent words • Some of the words we are looking at have a frequency which means we can fully exploit the annotation on the corpus with some confidence • Fuck is a good example of such a word • So let’s look at fuck

Male v. Female - word forms • Note that while quantity differs, ranking and proportions remain fairly stable. So while swearing may differ quantitatively, it does not differ qualitatively. Same is true of marked female words, like shit.

Full Categories

Categories type A B C D E F G I L M N O P R T M 53 202 57 70 1131 190 799 234 58 69 413 19 403 9 19 F Tendency 52 115 53 45 822 200 1250 225 90 44 517 20 359 13 35 M M M F F

Swearing in reported speech • Why is the relative proportion of reported uses of fuck higher for females (roughly one sixth of examples as opposed to one thirty fifth? )

Targets • Proportionately, more female uses of fuck are aimed at females than male uses of fuck, and more male uses of fuck are aimed at males than female uses of fuck. We seem to swear at our own sex most frequently.

Keywords • Notice we have evidence for a tentative explanation of the reported speech discrepancy

New Work – US Speech • Longman Corpus of Spoken American English (Du Bois for Longman) • Work undertaken with Neil Millar • Approximately 5, 000 words of orthographically transcribed spontaneous speech

female 3681 male 4585 unknown 3399

Word fuckage fucked fuckers Fuckery fuckheads fuckin Fucking fucks F 1 17 8 2 M 2 86 31 12 3 1 9 168 1 40 444 2

lemma female shit fuck god gosh damn hell gee/geez ass goodgrief/goodness/gracious/heavens 379 255 795 404 179 187 298 73 166 996 883 471 179 343 336 206 186 43 M M F F M M F piss bitch dang/darn shoot crap gay heck idiot jesus golly Lord jerk mother fucker asshole nigger 114 71 96 91 53 50 69 37 31 38 30 42 7 10 5 98 73 62 43 83 73 42 47 42 30 23 18 30 32 27 F M F F M M F F F M M M

Cat A B C D E F G I L M N O P T X F 25 71 18 2 223 299 2146 230 117 79 23 M 43 98 82 4 541 535 1608 356 185 99 26 UK Data M M M F F M F 151 274 17 6 610 352 21 25 F M F F

ab G E P N F I L A B D C M O T R c 1 G E N P I F B A L M C T D O R c 2 E G N P I F B C D M L T A O R de E G N P I B F D M C L T A O R Cat G F E O P I L M B C A N T X D Total 3754 834 761 626 586 302 178 169 100 68 49 38 31 6

Conclusion • Work on-going – the exploitation of the US data is far from being complete. New UK dataset available. • Similar patterns because of a shared cultural heritage? • Corpus data can, and has been, of use in the study of swearing. It is of particular use in looking at differences in usage through a range of variables • It is certainly an area where the corpus and other methodologies can combine