JournalofHistoricalSociolinguistics2016;2(2):155–188 Katherine McDonald* The sociolinguistics of gender, social status and masculinity in Aristophanes DOI10.1515/jhsl-2016-0011 Abstract:Thisarticleexploresvariationinthelanguageofmalecharactersinthe plays of the Athenian playwright Aristophanes, using Thesmophoriazusae and Frogs as in-depth case studies. Studies of modern languages have shown that men’slinguisticpracticescanbejustasmarkedforgenderaswomen’s,andthe data from these plays bears this out. Using past work on ‘female speech’ as a starting point, this article explores the incidence of gendered markers in male characters’ speech, and shows that some of these features characterise not just gender but the intersection of different aspects of identity including gender, social class and sexuality. These features include particular oaths, obscenities, certain uses of the particle ge, hedging and politeness strategies. The article showsthatalackofmale-associatedspeechmarkersisenoughtocharacterisea male Greek speaker as ‘unmanly’, without the addition of female-associated speech markers. Keywords:ancientsociolinguistics,AncientGreek,masculinity,genderlinguistics 1 Introduction Despiteongoinginterestinbothmasculinity(Gleason1995;FoxhallandSalmon 1998a, Foxhall and Salmon 1998b; Bassi 1998; Rosen and Sluiter 2003; Sutherland 2005; Goldhill 2004) and the language of women in the ancient world (Gilleland 1980; Adams 1984; Adams 2005; Bain 1984; Maltby 1985; Sommerstein 1995; McClure 1999; Mossman 2001; Schauwecker 2002; Willi 2003;Duhoux2004;Fögen2010;Clackson2011;Kruschwitz2012),thelanguage of ancient men and its relationship to the performance of masculinity has been neglected. To tackle this problem, this article turns to the plays of the fifth- century Athenian dramatist Aristophanes as one of our best sources for the linguistic performance of masculinity in the ancient world. As an initial exploration of the sociolinguistics of masculinity and its rela- tionshipswithgender,sexualityandclass,thisarticleinvestigatesthelanguage *Correspondingauthor:KatherineMcDonald,GonvilleandCaiusCollege,UniversityofExeter, Exeter,UK,E-mail:[email protected] Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM 156 KatherineMcDonald of male characters in two plays of Aristophanes: Thesmophoriazusae (c. 411 BC) and Frogs (405 BC).1 These plays, which will be used as in-depth case studies, havebeenchosentoprovidedataforthisstudybecausebothfeature‘unmanly’ men alongside more typically masculine charactersof a range of social classes, facilitating comparisons between different portrayals of male speech.2 Both plays feature Euripides, allowing us to explore whether Aristophanes gave the character (based on the famous dramatist) a similar linguistic characterisation acrosstwodifferentplays.Thesmophoriazusaeisalsofamouslyaplayinwhicha male character, Euripides’ Inlaw, tries to disguise himself as a woman, both in dressandinlinguistichabits.Thecontrastsbetweenthelinguisticcharacterisa- tionsofdifferentmalecharactersinthetwoplaysandthelinguisticbehaviourof a male character trying to conform to typical female language use in the Thesmophoriazusae will allow us to make some preliminary comments on how Aristophanes portrays male characters performing their gender and other aspectsoftheiridentitythroughtheiruseoflanguage.Thecharacters,situations and plots in the two plays are completely different from each other, and so the patterns of linguistic usage may not be directly comparable between the two plays;however,anypatternthatwedofindinbothplaysmaybeevidenceofa wider phenomenon worthy of further study. 2 Using Aristophanes as a linguistic source Aristophanes was an Athenian playwright who wrote a large number of come- diesforpublicperformanceinthelatefifthandearlyfourthcenturyBC.Eleven of his plays have survived to the present day. It has been recognised for some time that the plays of Aristophanes can give us some evidence for social varia- tion within Greek, including gender-based variation (Willi 2003; Duhoux 2004; Clackson 2011: 505–506).3 It has also been argued that the language of these 1 IamusingtheeditionsSommerstein(1994)andSommerstein(1996). 2 Iuse theterm‘unmanly’, asinRademaker(2003),ratherthan‘effeminate’,asIwouldnot wanttoassumethatmencanonlyfailtobesufficientlymasculinebybeingfeminine.Seealso Gaudio(1994). 3 Foramorescepticalview,seeBain(1984);Silk(1990);Dickey(1995:261–262);Colvin(1999: 286);seealsoPlutarchMor.853c-d.Silk’s(1990)argumentagainstthisuseofAristophanesasa sociolinguisticsource–thatAristophanicdialogueisinherentlyunrealisticbecausethestylistic levelofacharacter’sspeechalwayschangesduringtheplay–isnotasgreataproblemasit first appears.Firstly,a speaker’s or writer’s stylewill alwaysvary dependingon contentand context,andthishastobetakenintoaccountinanysociolinguisticstudy.Secondly,someof the extreme and non-naturalistic stylistic variation in Aristophanes – such as mock-tragic or Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM ThesociolinguisticsofgenderinAristophanes 157 playsmayreflectcontemporaryspokenlanguage,thoughtheevidenceneedsto be treated carefully. AfragmentofoneofAristophanes’lostplays(fragment706)hasoftenbeen quoted to show that he recognised social variation in language, including variation according to gender, social status and geographic location. The frag- ment reads: “… [him] whose language is the average style of the polis, neither urbane and slightly womanish nor vulgar and somewhat boorish” (trans. Willi 2003).4Whatisnotoftenstressed,however,isthatthisfragmentappearstodeal with variation within men’s speech, and not variation between two genders. It implies that women and men speak differently, but only indirectly, by stating that some men speak in a ‘womanish’ way – their language is somehow insufficiently masculine, or excessively feminine. What Aristophanes might mean by this, and how this insight might be evidenced in his work, is worth exploring. Past work on the sociolinguistics of Aristophanes has been invaluable in identifying possible markers of gendered speech in Greek. However, because of the wide range of gender identities and performances depicted in Greek drama, theuseof‘malespeech’asthestandardand‘femalespeech’asadeviationfrom the standard has left scholars needing to explain away the fact that some of Aristophanes’ characters do not conform to the typical speech patterns of their own gender.5 For example, it has already been noted that some female char- acters adopt aspects of ‘male’ speech styles when in positions of power (Willi 2003: 170; Taaffe 1993: 64) and that female characters violate gender norms around obscenity in certain circumstances (Willi 2003: 188), suggesting that thereisanimportantintersectionbetweenthelinguisticperformancesofgender and social status which needs further investigation. Scholars have also needed to explain away the avoidance of obscenity in the language of ‘unmanly’ male characterssuchasAgathoninThesmophoriazusae(Henderson1991:87;McClure 1999: 226; Willi 2003: 165; Fletcher 2012: 205 n. 8). There are also a number of occasions in Aristophanes’ plays when men impersonate women, or vice versa, and it is clear from the text that they make mock-legallanguage–issopronouncedthatitiseasytospotandaccountforinouranalysis. Silk’s argument does, however, remind us that we should take particular care with making generalisationsinonecharacter’sspeechacrosswholeplays,andthatcharacterswithshorter speakingpartsmayshowtheclearestandmostconsistentlinguisticcharacterisation. 4 διάλεκτον ἔχοντα μέσην πόλεως οὔτ’ ἀστείαν ὑποθηλυτέραν οὔτ’ ἀνελεύθερον ὑπαγροικοτέραν. 5 Scholarshaveusedboth‘gender’and‘sex’ascategorylabelsintheirdiscussionsofancient languagevariation.Iuse‘gender’heretorefertoasocialcategorywhichisperformedbythe individual. Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM 158 KatherineMcDonald use of linguistic as well as visual disguises (Taaffe 1993: 87–90; Sommerstein 1995: 63; Mossman 2001: 374; Stehle 2002: 387).6 But these disguises are not always straightforward: in Thesmophoriazusae, the typically masculine Inlaw uses feminine-sounding language even before he has been dressed up as a woman (see 3.4, below). The humour and playfulness around linguistic gender norms in Aristophanes means that even those linguistic features that are nor- mally labelled as ‘gender-exclusive’ are available to be used by anyone under appropriatecircumstances,howeverrareorcomicthosecircumstancesmightbe (Gilleland1980:182;Sommerstein1995:65–68;Willi2003:189–190;Fögen2010: 322–323).7 This kind of variation is difficult to explain while using ‘female speech’ as a monolithic category. Research into the sociolinguistics of gender in ancient languages therefore needs to take a new approach. The current focus on ‘women’s language’ and ‘femalespeech’misleadinglyimpliesthatmen’sspeechandwritingrepresentsa default or standard variety from which women deviate (Motschenbacher 2010: 1–2). Studies of both ancient and modern languages have already shown that there are markers associated with men’s language use as much as there are markers associated with women’s language use, particularly in stereotypical portrayals such as comedy. We also need to take note of developments in modern gender linguistics, including the growth of queer linguistics, which have shown that the gender binary is not the only possible starting point for linguistic research (Bucholtz and Hall 2004; Speer 2005: 12–14; Holmes 2007; Motschenbacher 2010: 7–14; Coates and Pichler 2011: 1–2, 370–373).8 Scholarship has started to explore variation within speakers of the same gender, rather than always designating data as ‘male’ or ‘female’. We must also recognise that there is considerable overlap between how people of different genders speak, and that there is wide variation within the speech of people of the same gender, including within the speech of one individual (Motschenbacher 2010: 20–25). With this in mind, I 6 The women of the Ekklesiazusae also have to practice making their speech sound more masculinebyavoidingself-referentialgrammaticallyfeminineformsandcertainphrasessuch as ‘by Aphrodite’ (Ekklesiazusae lines 155–195). In Thesmophoriazusae, the Inlaw tries to use feminine-soundinglanguagewhendisguisedasawoman(Thesmophoriazusaelines279–650), thoughitisdebatedwhetherornotheissuccessful. 7 Recent work on the performance practices of ancient comedy(Stehle 2002; Compton-Engle 2003)hasstressedthatallofthefemalecharacterswerevisiblyplayedbymen–sothatfemale speech in comedy can only be males imitating female speech. We cannot, strictly speaking, labelanyusageincomedy‘gender-exclusive’. 8 See also Gleason (1995) for the argument that Greek medical thought did not conceive of genderasbinary–instead,eachindividualwasamixtureofbothgenders. Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM ThesociolinguisticsofgenderinAristophanes 159 have avoided the term ‘female/male speech’, which might imply two clear and homogeneous categories. Thisarticlealsomakesuseoftheideaof‘doinggender’–thatis,theideaof gender as an ongoing continuous performance rather than a state that is achieved – and sees language as a key component in performing a gendered identity (West and Zimmerman 1987; Butler 1988; Butler 1990: 24–25; Butler 1997: 7–8; Gleason 1995: 70; Cameron 1997: 47–48; de Klerk [see Klerk] 1997; Sutherland 2005: 52; Holmes 2001; Holmes 2007: 51–52; Leap 2008: 283; Baker 2008: 72–89). However, it is rare for only one linguistic feature to be used to expressgender,orforanindividuallinguistic featureto expressonlygender.A single feature may index the intersection of a number of different identities, such as gender, social status, geographic origin and age, and the expression of an individual’s identity is achieved through a combination of interlocking features (Bucholtz and Hall 2004: 475; Cameron and Kulick 2003: 56–57; Colvin 2004: 95; Leap 2008: 283–284; Kiełkiewicz-Janowiak 2012: 318–319). This intersectionality is the reason why gender and social status must often be consideredtogether,andthefeaturesmarkingtheseidentitiescannotalwaysbe neatly separated. We should not always expect to find exact parallels between modern and ancient gender linguistics. The linguistic features which we might identify as gendered markers will of course be different, and the degree of gendered linguistic variation is likely to be different across societies, with some societies showing more restrictive gender roles and/or more fixed norms of linguistic practice for each gender. But nevertheless we should expect to find linguistic gender variation in ancient societies in much the same way as in modern societies, following the uniformitarian principle of linguistic variation (Conde- Silvestre and Hernández-Campoy 2012: 2). It has also been argued that ancient Athens and (some) modern societies are typologically similar patriarchal socie- ties in which both women and men are judged by their performance of their gender, including their linguistic performance (Willi 2003: 164–165). It is possi- ble, as we will see, that some of the gendered linguistic markers which we can identifyinGreekmayshowsimilaritieswithmoderngenderedvariation,atleast insofar as we can posit cross-culturally useful categories such as politeness or obscenity. Nevertheless, we should be careful not to transfer modern European categories of gendered language onto historical societies unless the evidence gives us good reason to do so (Kiełkiewicz-Janowiak 2012: 324). Inoureffortstogetasmuchaspossibleoutoftheavailabledataweshould not, of course, expect Aristophanic Greek to be able to tell us everything about how gendered language was used by speakers of fifth-century Attic Greek. We are dealing here with stereotypical and exaggerated depictions of speakers by Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM 160 KatherineMcDonald one educated male native speaker, and not data from a range of speakers. This doesnotmean,though,thatthesefeatureshadnorealityoutsidecomedy.Many of the features discussed may have existed in contemporary speech, and could have shown sociolinguistic variation which was picked up on by authors and audiences.9Recent workonthehighly stereotypicallanguageofmodernadver- tising suggests that written texts can also reinforce what correct ‘masculinity’ and ‘femininity’ sound like, and that reading or hearing such stereotypes can create real-life linguistic behaviour (Motschenbacher 2010: 52–58; Shibamoto 1987: 40–48). It is possible that comedy, which was seen by a large proportion ofthecitizenpopulation,couldhavecreatedasimilarpositivefeedbackloopof stereotypes. Nevertheless, there are many features which might have been important to speech variation between different men which are impossible to recover, including aspects not coded in writing such as pitch and intonation (Gaudio 1994). 3 Linguistic markers of gender Themostcompletelistofgenderedlinguisticfeatures,whichIwillbeusinghere asastartingpoint,isfoundinWilli’schapteronfemalespeech(2003:157–197). These gendered markers have generally been identified from ancient comments on male and female language and/or from statistical analysis of the speech of male and female characters in Aristophanes. Some but not all of these features have parallels in modern languages. Most of these features have been identified as either ‘male-preferential’ or ‘female-preferential’. A very small number of highly gender-preferential fea- tures, such as self-referential grammatically masculine/feminine forms and certain oaths, have been identified in previous work as ‘gender-exclusive’, although in certain circumstances including quoting and impersonation they may be used by a character of another gender. The goal here is to identify how Aristophanes uses both ‘male-preferential’ and ‘female-preferential’ features in his portrayal of male characters, and how his use of these features may vary with gender presentation or with factors such as social class which may have been associated with certain gender presentations in his audience’s minds. The features discussed here can be loosely grouped into the following categories, each of which requires a brief explanation. 9 On the‘highperformance’styleofpre-planned,publicperformancesandthisstyle’sexag- gerateduseofexistingsociolinguisticvariables,seeCoupland(2007:146–171). Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM ThesociolinguisticsofgenderinAristophanes 161 3.1 Metalinguistic commentary As noted above, gendered speech was a phenomenon that was commented on byanumberofancientauthors(Gilleland1980;Fögen2010).Fragment706isa particularly famous instance, but Aristophanes’ surviving plays also comment on male and female linguistic usage. In particular, the term laleĩn ‘to chatter’ (also laliá ‘chat’ and lálos ‘chatty’) is often used by characters in the plays to refertowomen’sspeech,andisoftenassociatedwithnegativeattitudestohow womenspeak,ortothefactthattheyspeakatall.10Occasionally,thetermlaleĩn isusedneutrally(tomean‘totalk’),orcanevenbeusedpositively.Theneutral useofthisverbcould alsobeviewedasaninnovation, sincetheverblégein‘to talk’ was later replaced by laleĩn (Willi 2003: 191); see also Section 3.7. 3.2 Grammatically masculine and feminine forms OneofthemostobviouswaysinwhichspeechisgenderedinGreekistheuseof masculine and feminine forms – for example, in pronouns, nouns, adjectives andparticiplesreferringtothespeaker,theinterlocutororathirdparty.Inthese plays, characters are sometimes misgendered by other speakers, while charac- ters who are temporarily impersonating a member of another gender may also refer to themselves with different forms than normal. 3.3 Oaths Theterm‘oath’refersheretoexpressionswithmá,nḗ,andprósplusthenameof a deity, meaning something like ‘by [god]; in the name of [god]’.11 The use of differentoathsbymenandwomenisoneofthemostobviousgenderdividesin spoken Greek. Some oaths are strongly gender-preferential, to the extent that unusualusagesmayattractcommentsfromothercharacters.Charactersofboth gendersswearbyZeus,andthisisthemostcommonoathamongbothgenders. Women also commonly swear by ‘the two goddesses’ (Demeter and Kore), Aphrodite, Artemis and Hecate; men generally swear by various male gods or by Demeter. Men may swear using ‘female’ oaths and vice versa for various reasonsofplot,suchasdeliberatelyimpersonatingamemberofanothergender, 10 Dover(1993:22)suggeststhatthetranslation‘chatter’or‘babble’canoftenbetoostrong, andthatitsmeaningiscloserto‘talk(toomuch)’or‘talk(outofturn)’. 11 ‘Oaths’inthiscontextdonotincludedirectaddressestogodsintheform‘O[god]’. Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM 162 KatherineMcDonald so that none of these oaths are truly gender-exclusive. Sommerstein also notes that Aristophanic women swear more often than men in general, and suggests that this makes their language sound more emotional (1995: 65–68). 3.4 Obscenity Henderson’s Maculate Muse gives the best list of Greek primary obscenities, that is,“wordsthatreferdirectly,withoutanyintermediaryassociationsordistancing, tothesexualorgans,excrement,andtheactswhichinvolvethem,andwhichare always improper” (1991: 35). These terms are: péos ‘cock’, kústhos ‘cunt’, psol̄ḗ ‘hard-on’,stúesthai‘haveahard-on’,skor̄ ‘shit(noun)’,khézō‘shit(verb)’,prok̄ tós ‘arse’,pérdesthai‘fart(verb)’(nounpordē)́,bineĩn‘fuck’,déphesthai‘wank’,kineĩn ‘fuck’, lek̄ añ ‘suck cock’, laikázein ‘suck cock’, splekouñ ‘fuck’.12 We should also include wordslike katapúgōn and eurúprok̄ tos ‘wide-arse’ i.e. a man who enjoys beingpenetrated.13Incontrast,thereareanumberofcommonlyusedeuphemistic or childish words, especially for body parts, which are less obscene: e.g. pugḗ ‘bum’,pósthē(ordiminutivepósthion)‘willy’,delphákionorkhoiríon‘pussy’(both lit. ‘little pig’).14 Some characters also make use of various metaphors for poten- tially obscene topics, some of which are probably off-the-cuff rather than estab- lished euphemisms(Adams1982: 3). Womeninmodernsocietiesareoftenthoughttousefewerobscenitiesthan men, though this is probably impressionistic or even prescriptive (Lakoff 1975: 50–51;Coates1993:126),andrecentresearchraisesseriousquestionsaboutthis generalization(deKlerk[seeKlerk]1997;Baker2008:48).Itispossible,assome studieshavesuggested(Hughes1992:294),thatbothmenandwomenusemore obscenitiesinsingle-sexgroups,andthusthatwomen’sfullrangeofusagehas historically been systematically hidden from male writers. It is, however, a persistentcross-culturalstereotypethatwomen,whoselanguagemaybepoliced 12 In general, I have followed the translations used by Sommerstein. These translations are intendedtoreflectapproximatelythecorrectlevel ofobscenity,usingBritishEnglishequiva- lents, though an exact reflection of the connotations of each word would be impossible. See Clackson(2015:Ch.5)fortheapparentstrengthofvariousancientobscenities. 13 Ontheconnectionbetweenunmasculinespeechandunmasculinesexualdesireimpliedby theseterms,seeWorman(2008:17).Davidson(2007:113)differsfromothercommentatorsin notseeingtheseinsultsassexual;AustinandOlson(2004:121)suggestthattheseinsultscanbe sexual in origin without referring to the sexual behaviour of the target (cf. ‘cocksucker’ or ‘wanker’asanall-purposeinsult). 14 Silk(1990:152)overestimatestheobscenityofkhoiríon,andmistakenlythinksitisout-of- characterforanoldwomantousethisword. Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM ThesociolinguisticsofgenderinAristophanes 163 more than men’s, use euphemism rather than obscenity (Coates 1993: 126; de Wit-Tak[seeWit-Tak]1968:363).Previousresearchsuggeststhatthisstereotype also applies in ancient societies, and it is understandable that ancient authors therefore characterize different characters’ speech by the use or non-use of obscenity (Willi 2003: 188).15 The use of obscenity may also relate to social status or age as much as gender. Occasionally in Aristophanes, an obscenity is greeted with surprise, shock or hostility from another character. This indicates the word has been used in a socially unacceptable way and may help us to interpret social attitudes to certain words, though there is no reason to think that the audience would have found obscenity shocking in itself (Robson 2006: 84). 3.5 Politeness Politeness includes linguistic strategies which relate to both the negative and positive face needs of the addressee (Brown and Levinson 1987). Negative face needs include the desire not to feel offended or insulted, and the desire to act without being obstructed by others. ‘Hedging’ is used here as a broad term for strategieswhichlimittheforceofaspeechact,includingwordsandphraseslike ‘maybe’, ‘you know’ and tag questions seeking agreement. These kinds of strategies address the negative face needs of the interlocutor. Hedging was originally identified as part of female linguistic practice by Lakoff (1975). More recent research, however, has shown that theuse ofhedging strategies by both men and women is complicated. Firstly, hedging does not always indicate hesitancy or lack of assertiveness, as Lakoff initially suggested: the same features can be used for expressing confidence, preventing talk from becomingtooface-threateningordrawinginotherspeakersasjointcreatorsofthe conversation (Coates 1993: 117). There is now considerable doubt over how far genderedhedgingstrategiesarefemale-preferential(FreedandGreenwood1996). Although we would not expect the specific politeness markers to be the same in Greek as in modern languages, it would not be unexpected for some social groups in any patriarchal or hierarchical society to use similar linguistic strategies. Some hedging and negative politeness strategies have already been identified elsewhere as linguistic features which may be a stereotypical way of representing some kinds of female or low-status speech. Features con- sideredhereinclude:doubleánwithinoneclause,hópōs+subjunctive+án,tag 15 PaceDover(1987:243):‘ItdoesnothappenthatsomecharactersinanAristophanicplayuse obscenelanguagewhileothersabstainfromit.’ Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM 164 KatherineMcDonald questions,and(morebroadly)unnecessarilyindirectorlitoticexpressions(Willi 2003: 177, 181–186).16 There are also politeness strategies which aim to protect the positive face needsoftheinterlocutor,suchasthedesiretobelikedandapprovedof.Thereisa general belief that women’s language usage includes more positive politeness strategies(Kramer1977:159;Holmes2001:159),thoughinpartthismighthaveto dowiththebehaviouroffemalesubjectsininterviewcontexts(Hughes1992:294; Holmes2001:162).Womenhavebeenfoundtousemorecomplimentsthanmen, particularly in all-female groups; men rarely compliment each other as a way of bonding.17Whilemendosometimescomplimentwomen,particularlywomenwho areclosetothem,theycanalsousecomplimentsinaface-threateningway–for example, by making overtly sexual comments that can be perceived as harass- ment (Coates1993:128–129;Holmes1995). ThefeaturesofAristophaniclanguagewhichhavealreadybeenidentifiedas positive politeness strategies used disproportionately by women and/or lower- class speakers include boúlei + subjunctive as a periphrasis for the deliberative subjunctive, the ethic dative, the fossilised imperative amélei ‘don’t worry’, and (morebroadly)complimentscateringtopositivefaceneeds(Willi2003:179–193). 3.6 Use of particles Greek particles can have many different pragmatic uses, from hedging and politeness to acting as intensifiers. The particle ge is an interesting case, as it can be used both as a hedge, by limiting a statement (‘I at least’), and as an intensifier (Willi 2003: 183; Denniston 1996: 114–115).18 But even as an 16 Theparticleán(whichisobligatoryinAtticinmainclauseswithacounterfactualindicative verborapotentialoptativeverb)cansometimesberepeatedinlongclauses.Willi(2003:181– 182) identifies this as a hedge, particularly in shorter clauses where the repetition cannot be motivatedbyneedingto‘remind’thelistenerofthisinformation. 17 ThecomplimentsbetweenthewomeninthefirstsceneofAristophanes’Lysistrataarecross- culturally interesting from this point of view, since this seems to show a group of women speaking and bonding unobserved by men, including complimenting strategies. However, as Fletcher(2012:226)notes,Calonice’s commentaboutLampito’s marvellous breasts“ishardly the type of greeting one woman might make to another”, and she argues that this is to be understood as a lewd comment from a (male) actor about another actor’s costume. See also Willi(2003:190);Robson(2006:49). 18 Ge issometimesreferredtoasa ‘scopeparticle’, whichdelineatestheapplicabilityof the utterancetoatleasttheitemgovernedbyge(Wakker1996:250).Itisnotalwayseasyoreven possibletogiveanaccurateaccountofthesemanticandpragmaticpurposeofaparticleina given context, and judgements tend to be somewhat impressionistic (Kroon 1995: 41–57, Brought to you by | Cambridge University Library Authenticated Download Date | 1/12/17 1:21 PM