Conversational Dynamics Of Humour

Journal of Pragmatics 35 (2003) 741–769 www.elsevier.com/locate/pragma

Conversational dynamics of humour: the telephone game in Greek Eleni Antonopoulou*, Maria Sifianou Faculty of English Studies, School of Philosophy, University of Athens, Panepistimioupoli Zografou, 157 84 Athens, Greece Received 20 December 2001; received in revised form 2 October 2002

Abstract The aim of this paper is to investigate humorous exchanges in Greek telephone conversation openings in the light of Raskin’s (Raskin, Victor, 1985. Semantic Mechanisms of Humor. D. Reidel, Dordrecht/Boston/Lancaster) and Attardo’s (Attardo, Salvatore, 1994. Linguistic Theories of Humor. Mouton de Gruyter, Berlin; Attardo, Salvatore, 2001. Humorous Texts: A Semantic and Pragmatic Analysis. Mouton de Gruyter, Berlin) semantico-pragmatic theories of humour and the principles of conversation analysis regarding telephone interaction [Sacks, Harvey, 1995. In: Jefferson, G. (Ed.), Lectures on Conversation, Vols I and II. Blackwell, Oxford (1963, 1970, 1972) (reprint) and Schegloff, Emanuel A., 1972. Sequencing in conversational openings. In: Gumperz, J.J. and Hymes, D. (Eds.), Directions in Sociolinguistics: The Ethnography of Communication. Holt, Rinehart and Winston, New York, pp. 346–380 (1968) (reprint)]. The material analysed (268 humorous exchanges between young adults) shows that such interactions are understood as a game, with interlocutors negotiating and co-constructing tacit rules involving a deliberate attack on social and linguistic conventions while at the same time creating a new code pertaining to in-group members only. The exchanges examined involve wordplay, insincere enquiries, complaints and reprimands. Wordplay in natural conversation has been attributed both an aggressive and a disruptive function (Norrick, Neal R., 1993. Conversational Joking. Indiana University Press, Bloomington, Indianapolis). Our data point to degrees of disruption, in that despite the playfulness of the exchanges, the canonical pattern including preemptive moves is preserved in most cases. Aggression, on the other hand, is also shown to be scalar and to serve primarily bonding purposes. In the light of the findings we propose a bridge between the GTVH, CA and politeness theory (Brown, Penelope and Stephen Levinson, 1987. Politeness: Some Universals in Language Usage. Cambridge University Press, Cambridge), with accompanying modifications considered necessary to account for this type of data and possibly for dialogic material of other types. # 2003 Elsevier Science B.V. All rights reserved. Keywords: Conversation analysis; General theory of verbal humor; Politeness; Social and linguistic conventions; Telephone call openings; Greek * Corresponding author. E-mail addresses: [email protected] (E. Antonopoulou), msifi[email protected] (M. Sifianou). 0378-2166/02/$ - see front matter # 2003 Elsevier Science B.V. All rights reserved. PII: S0378-2166(02)00150-9

742

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

1. Introduction This paper attempts to investigate humour in Greek telephone conversation openings between young adults. Such openings instantiate a very specific speech event governed by well-established linguistic and social conventions providing a framework against which interesting deviations can be studied. As our data reveal, verbal play is an essential aspect of telephone conversation openings between intimates with humorous utterances replacing turns and/or sequences of the canonical pattern (see Section 2.1). The most striking features in the data examined are playfulness but also aggression. The aggressive nature of humour has been extensively discussed in the literature, psychological, sociological and linguistic (see Attardo, 1993: 555). However, such research has focused on the function of aggressive jokes targeted at out-group members, thereby reinforcing the bonds between in-group members. Here reflexive aggression is exhibited by both interlocutors targeting each other and by extension rule governed behaviour in order to maintain intimacy (see Kotthoff, 1996; cf also Tannen and Kakava, 1992). The playful nature of the data is worth exploring within the framework of theories of humour, especially the General Theory of Verbal Humor (henceforth GTVH) (Attardo and Raskin, 1991; Attardo, 1994, 2001) (see Section 2.2). Playfulness exhibited in the data shows that the type of interaction we are examining is understood by our subjects as a specific game where both interlocutors follow tacit rules. These rules are co-constructed and probably negotiated at every exchange, involving a deliberate attack at standard conventions while at the same time leading to the creation of a new code pertaining to in-group members only. Culturally specific indirect mechanisms are eloquently discussed and exemplified in Kotthoff (1999: 135) where implicit norm negotiation is understood as ‘‘a metapragmatic function of humorous communication in the sense of an indirect indexing of shared values in orientation to each other’’. We, therefore, consider the material analysed illuminating as to the role of humour within the domain of in-group negotiation of social norms. We also consider the strategies employed interpretable in the light of using discourse for bonding purposes (cf Attardo, 1993: 556), an obvious priority for any social group which is positive politeness oriented (see Sifianou, 1992). We are interested in exploring: (1) how the humour in these instances is best analysed and (2) what specific sociocultural purposes it serves in this context. The data analysed consist of 121 telephone call openings recorded with the aid of a small cassette player attached to the plug of five different people’s telephone. Thus both in-coming and out-going calls were recorded. In addition, another 675 instances were analysed from data collected by students who were asked to record their own telephone openings (both self-initiated and received) and transcribe them as accurately as possible on a specially prepared observation sheet. All subjects collecting the data filled in contextual information, such as the gender, approximate age, status, and degree of familiarity between interlocutors. Telephone interactions between strangers were excluded from the sample (796 exchanges in all). Our corpus includes exchanges primarily between conversationalists having ‘‘personal’’ and

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

743

‘‘familiar’’ symmetrical relationships (see Pavlidou, 1994: 491 for this classification). Of these, 268 included some kind of humorous utterance. The specific strategies used may largely reflect the disposition and sociocultural conditions of our subjects (mainly Greek university students from various regional backgrounds). We will first present briefly the institutional frame for telephone conversation openings in terms of Conversation Analysis (henceforth CA) and then discuss a possible application of the GTVH to such conversational material. We are in full agreement with Attardo (1994: 293) that ‘‘conversation analysis is the field of linguistics best suited for investigating humor in its spontaneous setting’’, which should, however, incorporate social aspects of context (see Mey, 1993: 185). A detailed analysis of the data and a final discussion of the relevant theoretical issues will follow.

2. Theoretical background 2.1. Telephone conversation openings Work on opening telephone interactions pioneered by Sacks ([1963] 1995) and expanded by Schegloff ([1968] 1972) identifies a set of four core sequences typical of such interactions in North America. These are: (1) a summons/answer sequence, (2) an identification and/or recognition sequence, (3) a greeting sequence and (4) an exchange of ‘howareyou’ sequences. This pattern has been characterised as canonical and is assumed to be universal. In relation to Greece, this canonical pattern is only evidenced between conversationalists whose relationship is distant, either vertically or horizontally, and even in such cases an interlocking rather than a serial organisation seems to be preferred.1 Between closely related interlocutors, the actual canonical pattern, that is, the most frequently followed one, normally involves only two sequences, those of summons/answer and ‘howareyous’, through which mutual recognition is achieved (Sifianou, 2002). This pattern reflects the fact that in Greek both callers and answerers expect other recognition and avoid explicit self-identification, as earlier research has shown (Bakakou-Orfanou, 1988–1989; Pavlidou, 1994; Sifianou, 1989). Failure of recognition seems to be interpreted as name forgetting (see Brown and Levinson, 1987: 37). In addition, greetings are not as frequent in Greek as are initial enquiries. In some cases, before the opening is fully worked out, one of the interlocutors may introduce the first topic in a ‘‘preemptive move’’ (Schegloff, 1986: 133). Of special interest to us in this paper are preemptive moves involving complaints (ibid.: 144). These are usually contact related and are performed by answerers, while for callers preemption usually involves extrinsic matters of urgency. In relation to the language used in Greek telephone call openings, a further issue of relevance is the rather personal style for answering a home telephone, given the 1

Serial organisation means that there are four clearly identifiable sequences whereas in interlocking organisation, some turns have two or three components (Schegloff, 1986: 130–131).

744

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

variety of linguistic expressions available to respond to the summons. Some of these are: l"
"" ‘speak.PL’, l!
‘please’,  or 
l  ‘yes’, "o´& ‘go ahead’, oı´ " ‘order.IMP’, the infrequent "
l ‘come’,2 and even combinations of these like l"
"" l!
‘speak please’ or o o& "ı´  l!
‘who is it please?’3 As in all languages, what the answerer essentially indicates is that contact has been established so the caller can go ahead and speak. From this variety of utterances, speakers typically adopt one for regular use thus providing a ‘‘signature’’ similar to the ‘‘signature hello’’, that is, ‘‘a distinctive mode of delivery, more or less standardised across occasions, which provides for ready recognition’’ (Schegloff, 1979, 1986: 123). In addition to the above organisational and linguistic conventions, there are social conventions governing telephone calls. The primary function of the telephone in Greece seems to be interactional (rather than transactional), especially among young and, interestingly, among elderly people. The Greeks seem to have endorsed the British Telecom advertising slogan ‘‘It’s good to talk’’. To call friends and relatives is a social need or an obligation which has to be obeyed. Failure to make such calls can be seen as lack of interest or concern and even as aloofness and snobbery which may result in negative judgements, complaints and even sanctions. The more-or-less obligatory regular telephone contact to simply exchange news or chat among friends, especially in big cities, seems to have replaced the casual, unexpected visits people used to pay in the past. A telephone call is more immediate and thus preferred to sending letters or cards on many social occasions. Expectedly, such attitudes lead to frequent and lengthy calls causing dismay to those who might be trying to get in touch in the meantime. Consequently, complaints for either prolonged lack of contact or busy lines are not infrequent [see examples (12)–(15)] and reflect a kind of friendly sanctions, thus constituting legitimate sequences of the opening. These are often expressed indirectly and playfully, as we will see below. 2.2. Theories of humour and conversational data Linguistic theories, at least in their initial stages, are strongly influenced by the type of data they are first formed to account for. A standard example is reliance on a single language, for example, English (or in fact American English) for the best part of the last century. Neither Generative Grammar nor pragmatic theories escaped from this pitfall as they were initially conceived. At its inception, CA was also bound to the American cultural context. Humour theory seems to be modelled on the type of text it set out to cater for, i.e. the joke. Conversational humour, insightfully discussed in the work of Norrick (1993) and Kotthoff (1996, 1999), for instance, has had to draw on theories not focusing on humour, and has been largely descriptive. 2
l " is the imperative of the verb "
o ‘come’ and is used in the opening sequences of informal telephone interactions between closely related interlocutors. See Pavlidou (1994: 495) and more specifically Pavlidou (1995) where the variety of functions of this lexical item are explored. 3 The history of these opening gambits is far from clear and their rendition into English very difficult. It is noteworthy, however, that similar expressions are also found in other languages, like pronto ‘ready [to take your call]’ in Italian, prosı´m ‘I beg [the favour of your call]’ in Czech, diga ‘say [your message]’ or the more old-fashioned mande ‘command [me to answer you]’ in Spanish (Mey, 1993: 228).

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

745

The GTVH is, in our view, the only full-fledged semantico-pragmatic theory of humour today which is coherent, formalizable and epistemologically interesting. It expands Raskin’s (1985) model to include information beyond the semantic level and account for humorous texts of any length or genre (Attardo, 2001). This model has never been applied to text types similar to the ones under investigation, but we consider that (a) our data constitute a good case study for testing its validity and (b) the present analysis could contribute to useful modifications of the theory. For present purposes the following premises will suffice: 1. For a text to be funny it must be (fully or in part) compatible with two scripts opposed to each other. 2. Humorous texts adhere to a special kind of co-operative principle consisting of maxims jointly amounting to the Non-Bona-Fide (NFB) mode of communication, where speakers are not committed to the truth of what they say. 3. Humorous texts are informed by the following 6 basic parameters or Knowledge Resources (KRs): 1. Script Opposition (SO) (see 1 above). 2. Logical Mechanism (LM): the way in which the two scripts are brought together. 3. Situation (SI): objects, participants, activities, etc. i.e. the context. 4. Narrative Strategy (NS): narrative organisation of the text, including adjacency pairs and figures of speech. 5. Target (TA): the ‘‘butt’’ of a joke. 6. Language (LA): information necessary for the verbalisation of the text.

The GTVH includes additional insights from pragmatics, text linguistics, and theory of narrativity, although its semantic origin (i.e. Raskin, 1985) is still prominent, in our view. The boldest attempt at applying the model to material similar to ours is an analysis of a TV sitcom episode (Attardo, 2001: 128–134). The differences between sitcoms and telephone call openings are too obvious to require mention. Yet, the dialogic nature of both types distinguishes them significantly from both jokes and narratives. Notice first, that unlike in jokes, where the boundaries of the text are clear, in a dialogue, the unit of analysis is not immediately identifiable. Every turn typically serves as the trigger for the next one and by consequence, in playful dialogue, a chain of humorous utterances is created, with jab and punch lines interacting with each other and with bona fide material to co-construct the text. It is, in fact, unclear whether jab or punch lines are the objects of investigation. In Attardo (2001: 82–83), jab lines are defined in opposition to punch lines mainly on the basis of structural criteria: the punch line (Attardo, 1994, passim, 2001: 83) is final in jokes, while the jab occurs ‘‘in the body of the text’’, i.e. in any non-final position. Semantically, the two concepts are assumed to be identical. Textually, they are supposed to differ not only in terms of position but also in terms of function: the jab is non-disrupting the

746

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

narrative while the punch is identifiable as the disjunctor, i.e. the element which forces the reader to switch to the second script. In other words, the punch has a disruptive function by definition (and in accordance with its structural value in the text). However, if the text analysed is an adjacency pair, both turns of which include funny instances, it seems arbitrary to decide that the second one qualifies as a punch, simply because of the position it occupies. Similarly, if the text consists of a series of adjacency pairs, playful turns may be interrupted by bona fide ones before the playful key is resumed. Here again, it seems arbitrary to consider the last playful utterance as a punch line unless the contribution of the functional criterion of disruption is clearly and independently specifiable. We suggest that the structural/ positional difference, which, as already pointed out, presupposes a clear segmentation of the text of analysis in a non ad-hoc manner, should be considered in playful dialogue in relation to the degree of disruption effected by the humorous line under analysis (this will be discussed below). Two examples from our data will suffice to illustrate the issue:

(1)

C: ring A: nai; C: io
rgo; A: e
la C: hi [sic] A: wou
i C: kala
eı´sai; A: mia wara
. Esu
; C: ring A: yes? C: George? A: come on C: hi [sic] A: hui (hi) C: are you well? [how’re you?] A: very well [and] you?

The playful adjacency pair ‘‘hi’’/‘‘hui’’ [discussed below in detail as (11) and in footnote 16] takes up the slot of the greeting sequence, and is followed by a ‘‘howareyou’’ sequence as predicted by CA. In this sense, it is difficult to detect a disrupting function in either of these turns. Therefore, although structurally ‘‘hui’’ could be considered a punch line (as occurring in the second turn of the relevant adjacency pair, followed by a bona fide sequence), we will consider it a jab line as it does not exhibit a clearly disruptive function. In the case of (2),

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

(2)

747

C: ring A: nai; C: o´wi A: a! esu
eı´sai; ti ‘o´wi’ le& paida
ki mou; C: e
la re, ti ka
nei&; ti ne
a; A: kala
, esu
e
wei& e
xodo; C: ring A: yes? C: no A: ah, it’s you? what ‘no’ are you saying, my child? C: come on re.part,4 how are you doing? What’s new? A: fine and you? are you on furlough (off-duty)?

the playful utterance occurs in the first turn of the second adjacency pair (occupying the identification/recognition slot). The second turn of the pair verifies that identification has actually been achieved (probably through voice recognition) and is followed by the expected ‘‘howareyou’’ sequence. It seems therefore counter-intuitive to assign a disruptive function to it, simply because of the occurrence of the second part in A’s response: ‘‘what ‘no’ are you saying’’. In short, the canonical pattern is adhered to in both (1) and (2) so ‘‘hui’’ and ‘‘no’’ should be considered jabs as they are hardly disruptive.5

3. Challenging linguistic conventions The telephone game can start at the very initial stage, the summons response. Despite the variety of available options in Greek, some people use loan words, such as ‘‘pronto’’ and ‘‘hello’’ or even totally idiosyncratic expressions such as "
 "; ‘shoot’/‘fire away’, a loan from the football ground, ‘‘talk to me’’ or ‘‘yellow’’

4 Both " and " [in example (12)] are related untranslatable particles, they are, therefore, inserted transliterated in the literal English translation. They frequently accompany terms of address, reinforcing their positive or negative load although deriving from the Classical Greek adjective !o´& ‘stupid’. They are very frequent in everyday conversations and mark the context as informal (see Makri-Tsilipakou, 2001; Tannen and Kakava, 1992, and footnote 19). 5 Attardo (2001: 81) does address the problem of identifying the end of a narrative as a prerequisite for identifying the punch line (since end-point is its defining feature) and proposes possible empirical techniques for the segmentation of the text vector, for material of a completely different kind from the data under investigation. His actual analysis of the dialogic text in the case studies (ibid.: 128–134) shows that he identifies as punch (rather than jab) lines (1) self-contained jokes (followed by serious sequences) and (2) humorous utterances at the end of a scene (i.e. humorous utterances formally, structurally marked as final). Such indicators are absent in our data.

748

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

echoing ‘‘hello’’ from a well-known joke.6 Apparently, when answerers expect a call from a close friend, they may consider it appropriate to indicate their readiness to play from the very beginning. Rarely do callers ignore this invitation, so they tend to react equally playfully. Interlocutors may be in a ‘‘joking relationship’’ and this is reflected in the ways they enter their conversations (see Sacks, [1970] 1995: 206) because as Norrick (1993: 6) suggests, ‘‘these histories of joking have much relevance for any interactions of the people involved’’. Examples (3) and (4) are illustrative: (3) C: ring A: talk to me! [sic] C: ti ‘talk to me’ eı´n’ auta
; ego den spik inglis. C: ring A: talk to me! [sic] C: What ‘talk to me’ are these? I don’t speak English. (4) C: ring A: pronto C: e
la re, ti ‘pronto’ eı´n’ auto´; ma& to paı´zei& Italo´&; C: ring A: pronto C: come on re.part what ‘pronto’ is this? Are you trying to pass for an Italian? In (3) the caller comments on the inappropriateness of the answerer’s expression by playfully pretending rudimentary knowledge of English, and also adopting a funny accent. In (4), the caller recognises his friend’s voice (despite the unusual expression used), as well as the origin of the foreign word and teases him for attempting to adopt a foreign identity. So ‘pronto’ activates a script, roughly describable as ‘a telephone call response to summons by an Italian/to an Italian/in Italy’. In the absence of a pre-existing linguistic environment to activate the second, opposed script (to the effect that there are no Italians but Greeks involved and the event is taking place in Greece), this situational script is not activated by a linguistic unit as such.7 However, in accord with Schegloff’s ([1968] 1972: 357) suggestion, the telephone ring should be regarded as a summons parallel to a linguistically encoded one, which can be seen as activating the situational script described above.8 Given 6 The joke runs as follows: In an EFL class for children, the teacher is asking them to produce utterances using the words ‘‘green’’, ‘‘yellow’’, and ‘‘pink’’. A number of pupils perform as expected until Bobos (a stereotype for a naughty pupil) comes up with the following: ‘‘I’m doing my homework when suddenly the phone rings: ‘green, green!’, I pick it up and say ‘yellow!’ but I get no answer so I hang up: ‘pink!’’’. 7 Unlike the case of psychological theories and AI where scripts are understood as merely experiential/ cognitive objects, the GTVH assumes scripts to be evoked by linguistic units (Attardo, 1994: 200). 8 It may prove to be the case that for conversational humour a single lexical unit activates (simultaneously) the two (partially overlapping and) non-congruous scripts. This is a separate issue, for which data other than those under investigation can be illuminating.

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

749

this proviso, the SO can be easily identified as ‘normal/abnormal’ and ‘Greek/Italian’.9 The same applies to example (3), where the additional SO specification has to be not only ‘Greek/English’ but also ‘conventional response to summons/order to speak’. Even when the answerer uses one of the conventional options to respond to the summons, the caller does not miss the opportunity to initiate play, as in the following examples:

(5)

C: ring A: parakalo
; C: parakaleı´& 10 polu
; C: ring A: please? C: do you beg a lot?

(6)

C: ring A: parakalo
; C: mZn parakala
& kai polu
C: ring A: please? C: don’t beg too much

(7)

C: ring A: parakalo
; C: poion parakaleı´& kai giatı´ ton parakaleı´&; C: ring A: please? C: who are you begging and for what?

The expression l!
‘please’, is a verb meaning ‘to request politely’ or ‘beg’, also used in response to summonses, thanks and apologies.11 For the explanation of the above examples, we consider Giora’s (1997: 185) ‘graded salience hypothesis’ 9

For Raskin (1985: 113–114), SOs are first characterised as ‘normal/abnormal’, ‘actual/non-actual’ and ‘possible/impossible’ and further instantiated in more concrete ‘mid-level’ ones: ‘good/bad’, ‘life/ death’, ‘obscene/non-obscene’, ‘money/no-money’, ‘high/low stature’. Once again it seems to be the case that the joke orientation of the theory restricts the possibilities of SO to standard options of referential humour in standard jokes. Different types of humorous utterances will necessitate different oppositions at mid-level. 10 Pl"ı´& and its socially determined variant l
& in (6) are both second person singular forms of the verb l!
‘request politely’.

750

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

(GSH) most illuminating;12 here, salience is understood as a function of four parameters: conventional meaning, familiarity, frequency, and ‘‘givenness status’’. In the context under consideration, l!
‘please’ as issued in response to summons is salient as ranking high on all four parameters (with the notable exception of ‘frequency’) in the canonical pattern. By recontextualising it, the caller issuing the humorous utterance opts for a less salient interpretation, thus refusing to recognise its appropriateness (i.e. discarding the contextually most salient interpretation). In this way, the caller presents him/herself as ‘‘someone willing to suspend the conversational business at hand for a laugh’’ (Norrick, 1993: 60). The humorous effect created here aims at attacking conventional use as ‘too formal’, or appropriate for out-group members only. In other examples, the reason for recontextualisation is overt. Compared to alternatives like  ‘yes’ or the more colloquial "
l ‘come’, l!
‘please’ is perceived as more formal (along with l"
"" ‘speak-pl’ or the more old-fashioned oı´ " ‘order-imp and "o´& ‘go ahead’). In the following examples callers appear to be challenging the appropriateness of the response in terms of register, since the answerer is a close friend:

(8)

C: ring A: parakalo
; C: ti parakaleı´& re, o ana
sZ& 13 eı´mai. C: ring A: please? C: what are you begging re.part, I am Thanasis.

The criticism becomes explicit in the caller’s reaction in example (9), where the caller assumes  ‘yes’ to be a more appropriate response and makes a playful metalinguistic comment challenging the answerer as if she were a child being admonished for inappropriate verbal behaviour.

11

In response to thanks/apologies it means ‘‘I beg of you not to mention your gratitude/debt’’. The definitions of lo
are: 1. asking politely, begging (of somebody to do something); 2. in different expressions (often accompanied by you-SING/PL) through which something is requested politely, permission and orders are given. 3. as a positive response to a question, request or thanks. (Similar to prego, je vous en prie. . ..). Evidently, whether the particular use in question belongs to 1 (asking politely) or 3 (conventionalized response to request/summons), the senses are not dissimilar (Leksiko tis Kinis Neoellinikis, p. 1020). 12 GSH was developed for and addressed a different issue, namely salient meanings in relation to priority in interpretation. 13 Names and surnames in these examples are replaced by fictitious ones.

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

751

(9) C: ring A: parakalo
; C: de le
ne korı´tsi mou ‘parakalo
’. ‘Nai’ le
ne o´tan apantou
n sto tZle
fono. e sto e
woun ma
yei; A: Miwa
lZ, a
se tZ dia
lexZ
. erime
no tZle
fono apo´ to dieuyuntZ
tou swoleı´ou, gi auto´ m’ e
wei fa
ei Z euge
neia. C: Kala
, pa
o pa
so. C: ring A: please? C: they don’t say ‘please’ my girl. They say ‘yes’ when they answer the phone. Haven’t they taught you that? A: Michali, stop lecturing me. I’m expecting a call from the head master, that’s why I’m being troubled with politeness. C: Well, pass.

The answerer justifies her opting for a formal variant and the caller accepts the answerer’s justification. In terms of Script Opposition (SO), ‘normal/abnormal’ is to be supplemented with ‘responding to summons/requesting politely’ in all these cases, while in (9) where the speaker pretends to be admonishing the addressee for conversational norms, ‘adult to adult’ is opposed to ‘adult to child’. Needless to say that none of the mid-level oppositions are exhibited in this data. Interestingly, the Logical Mechanism Knowledge Resource (LM KR) here can be specified as ‘‘ignoring the obvious’’ (see Attardo, 2001: 27). The Language Knowledge Resource (LA KR) will accommodate information such as pun and metalinguistic comment [for (9)]. Inappropriateness in terms of register is obviously related to the degree of intimacy between interlocutors. Interestingly, it is as if callers assume that answerers should know that a friend is calling, so they should opt for less formal alternatives.14 This assumption may reflect appreciation of the telephone as a device for maintaining social contact and intimacy rather than conducting business where a more formal code would be required. On the surface, the Target Knowledge Resource (TA KR) is the addressee but in essence the attack aims at the conventions used. However, callers are not deterred from reacting playfully, even when the response is the perfectly informal and most frequent  ‘yes’ as in (2) repeated here as (10) for ease of reference: 14 Similar reactions are triggered by summons responses with l"
"" ‘speak’, the second person plural of the verb le
! ‘say’, also used in contexts of formality to address a single addressee.

C: ring A: le
gete; ‘speak-PL’ C: ti le
gete; ‘what ‘speak-PL’?’ Appropriateness of register is a rich source of humour in telephone call openings between young closely related interlocutors, to be discussed later in relation to failure of recognition (Section 4.2).

752

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

(10) C: ring A: nai; C: o´wi A: a! esu
eı´sai; ti ‘o´wi’ le& paida
ki mou; C: e
la re, ti ka
nei&; ti ne
a; A: kala
, esu
; e
wei& e
xodo; C: ring A: yes? C: no A: ah, it’s you? what ‘no’ are you saying, my child? C: come on re.part, how are you doing? What’s new? A: fine and you? are you on furlough (off-duty)?

As it transpires from the intonation pattern of the utterance, the implied message in the fourth turn is ‘‘how can you be so stupid’’ reinforced by the diminutive form ( 
 ‘child-dim’) which sounds belittling in this context. However, the next adjacency pair resumes the expected tenor in the exchange of ‘howareyous’, verifying the assumption that no negative evaluation was in fact intended by either interlocutor. Interestingly, it is the third turn which is genuinely incongruous here and could be considered a punch line, on these grounds. However, the fourth one is also playful as evidenced by its intonation pattern, rather than the actual lexical items used. Regardless of whether the GTVH could accommodate suprasegmentally provided information, this fourth turn, although final [i.e. occurring before the ‘bona fide’ (BF) mode is resumed] it is certainly less striking, funny or incongruous than the preceding one. It might, therefore, be useful to include considerations of relative incongruity in decisions on the discrimination between jab and punch lines in dialogue. Although the high salience of  ‘yes’ in these contexts is indisputable (see, e.g. Sifianou, 1989: 530), callers often invent humorous devices to challenge its use, echoing, for instance, the answerer’s response to the extent of imitating the tone of voice and the intonation originally used, or responding to it with an alternative summons response, like l!
‘please’. Thus what seems to underlie all these reactions is an attack on conventional summons responses, irrespective of actual linguistic realisation, as a signal of playful disposition.15 The SO in (10) is probably ‘actual/non-actual’, ‘responding to summons/answering a question’. The interest here lies in interpreting  ‘yes’ as a(n) (positive) answer to a question rather than as a response to summons, providing its antonym in the humorous utterance. Therefore, once again the speaker selects the meaning which is least salient in the given context, as with the l!
‘please’ examples. Similar defiance of conventions is evidenced in playful greetings, as in (1), repeated here as (11): 15

Such reactions to summons and summons responses are reminiscent of Schegloff’s (1979: 39) ‘‘joke first’’ strategy. See also Norrick (1993: 21–25).

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

(11)

753

C: ring A: nai; C: io
rgo; A: e
la C: hi [sic] A: wou
i C: kala
eı´sai; A: mia wara
. Esu
; C: ring A: yes? C: George? A: come on C: hi [sic] A: hui (hi) C: are you well? [how are you?] A: very well [and] you?

In this instance, after mutual recognition has been achieved, the caller uses hi to greet his interlocutor who responds with the totally irrelevant word o
‘habit’.16 In all probability o
is not intended as a word but as a nonsense sequence alliteratively connected to the playful code switching ‘hi’. The SO can be specified as ‘normal/abnormal’, ‘sense/nonsense’. Alliteration and ‘non-word’ will appear under LA. The TA, however, may well be ‘attack on the use of loan words’ if there is any TA at all.17 Otherwise, we are still dealing with a possibly culture-bound phenomenon of ‘abuse of verbal humour/abuse of language for the hell of it’. It is perhaps revealing in this connection that this particular joke triggered more laughter from Greek audiences than any other in the data.18

4. Challenging distance Although the instances analysed so far are not openly confrontational, one could discern covert aggression, especially if we accept Norrick’s (1993: 60) suggestion that ‘‘puns rank high on the scale of aggression’’. However, aggression under cover of a playful key is most obvious in cases where the answerer has not been in touch with the caller as frequently as, or at the moment that, the caller wishes contact to be established. Here the target is not a linguistic convention but the answerer’s actual 16

o
is a colloquial word meaning habit, usually strange or bad. Alternatively, the TA of the whole pair may be the norm indicating that greetings are not as necessary in Greek as they are in other languages (see Sifianou, 2002). 18 We have checked our data with other native speakers of Greek to assess that they sound as funny as we thought them to do. 17

754

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

social behaviour towards a close friend. This is how we suggest that the following example, among others, is to be interpreted:

(12)

C: ring A: nai; C: ti ‘nai’ bre mala
ka; me poion mila
& to´sZn o
ra; A: ‘nai’ re mala
ka. umZ
yZke& na ma& pa
rei& tZle
fono; C: a
se re to
ra, ta dika
sou. Ti gı´netai; kala
; A: ta ı´dia. Esu
kala
eı´sai; C: ring A: yes? C: what ‘yes’ vre.part you jerk? With whom have you been talking for so long? A: ‘yes’ re.part you jerk. You have remembered to call us? C: drop it re.part now, yours (news). Is everything well? A: the same. You, are you well?

The answerer’s response indicating readiness for communication is challenged by the caller, pretending to be angry for having had to cope with a busy line for long. This caller feels entitled to use the swear word l
& ‘jerk’ along with the particle ".19 Aggression is further assumed in the same turn, with the caller’s enquiry about the identity of the previous interlocutor, which is promptly interpreted as a rhetorical question and therefore receives no answer. Instead, the answerer repeats  ‘yes’ and uses the same term of address and a similar structure to challenge his friend for infrequent contact. To signal the distance he attributes to the caller, he uses the 1st person plural pronoun (& ‘us’) inappropriately, probably transferring the plural marker of formal politeness from the addressee (you) to the speaker (us). Laughter follows and the caller who initiated the game switches to ‘howareyous’. Aggressive utterances are interestingly reinforced with overstatements as in the following example:

(13)

19

C: A: C: C:

ring le
gete; e
la re mala
ka se ca
wno to´se& me
re&. Ti le
gete kai xele
gete; ... s’ e
cawna wye& kai prowye
&. ou brZ
ka douleia
.

C: A:

ring speak-pl?

The word l
ka& literally means masturbator but it is commonly used typically between young males but also some females to indicate in-groupness and intimacy. See also footnote 4 on ".

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

C: C:

755

come on re.part you jerk. I’ve been looking for you for so many days. ... I was looking for you yesterday and the day before. I’ve found a job for you.

As it transpires from the caller’s second turn, o
"& "
"& ‘so many days’ is merely ‘‘yesterday’’ and the day before. The intonation pattern used in the caller’s first utterance signals indignation, but it is really in the second component of the same turn that the playfulness of the first one becomes apparent. In this case, the ironic echo of the summons response is further marked through an additional repetition prefixed with ", a bound morpheme (much like the English ‘‘un’’) usually suggesting a reversal of the action expressed in the main verb, indignation or indifference.20 Here it is evidently pure wordplay with the implication ‘‘This is an urgent matter and you (being an asshole) respond to the summons in a formal, cool and neutral manner’’. In other words, indignation is rampant throughout this exchange, but the playful challenge to the response follows the first head-on attack, thus mitigating the aggressive tone of the first utterance (despite being indignant itself). Aggression and overstatement can merge in a single utterance. In another example, this combination reaches the extremes of the caller’s saying  " o!
! ‘I’ll kill you’ followed by the prospective murderer’s motive " ´ ! o´ o !ı´ ‘I’ve been looking for you all morning’.21 Exaggeration and overstatement are strongly preferred to understatement or litotes. Traditional figures of speech are accommodated under the NS KR in Attardo (2001, passim). Therefore, ‘overstatement’ should appear under NS. Exaggeration may also be a culturespecific tendency, with overstatement understood as the linguistic expression of overreaction. Under SO, the ‘normal/abnormal’ opposition can be further specified as ‘trivial/non-trivial’, i.e. treating trivial matters as if they were of the utmost importance. Aggressive humorous utterances require obviously a specification of the TA KR. On the surface, what is targeted is the particular addressee. In essence, however, the target is a type of social behaviour. 4.1. Being indirect through insincere enquiries Interestingly, all complaints in our data are performed indirectly through off record strategies, mainly enquiries (see Brown and Levinson, 1987: 221, 223), as in (12) above. This type of utterance (‘‘who have you been talking to for so long?’’) appears frequently in our data but is rarely answered, unlike in the following example (14):

20 This is an extremely productive mechanism in that " can practically prefix any word to challenge its information content or contextual appropriateness. 21 On the dissociation between words and actions in Greek, see Hirschon (1992).

756

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

(14) C: A: C: A: C: A:

ring poio& eı´nai parakalo
; e
la . . . Ypourgeı´o ekeı´; pou
mila
te to´sZn o
ra; e
la mpampa
, esu
eı´sai; milou
sa me mia fı´lZ mou. kala
kai ti le
gate to´sZn o
ra; to Kupriako´ lu
nate; perı´pou. 0Asto to
ra auto´. ia pe& mou ti ka
nei&;

C: A: C: A: C:

ring who is it please? come on . . .Ministry there? Where [sic] have you been talking for so long? hi dad, is that you? I’ve been talking to a friend. OK and what have you been talking about for so long? Have you been solving the Cyprus issue? A: just about. Drop it now. Just tell me how you are.

Here the caller camouflages his frustration for having been unable to get through to his daughter for some time, enquiring whether he has actually got through to a Ministry (where lines are constantly engaged). This is followed by another enquiry, rather incongruous for the assumed Ministry context, concerning the identity of the addressee’s previous interlocutor. Thus, the caller deliberately mixes up two distinct scripts exhibiting the opposition ‘father to daughter’/‘member of public to civil servant’ with  l
"; ‘speak’ (2nd person plural marking formality) as the linguistic signal of the confusion. The daughter in this case does provide a response (probably as an expression of respect for her father), but apparently does not address the issue implied in the father’s question, i.e. the length of the call. As a consequence, in his next turn, the father further enquires playfully whether she has been trying to solve the Cyprus issue with her friend, a good reason for a lengthy call. The absurdity of the suggestion (discussing a serious political issue instead of simply chatting with a friend on the phone) points, once again, to the opposition of two distinct scripts. A similar case is reflected in (15):

(15)

C: ring A: nai; C: geia sa& A: geia sa& C: ya Z
yela na ka
no mia aı´tZsZ ston [OTE A: [Sorry! [sic] C: gia na pia
so grammZ
A: e
la! C: kala
po´sZ o
ra mila
&; A: e de mı´laga kai pa
ra polu
C: xe
rei& po´sZ o
ra mila
&; C: ring

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

757

A: yes? C: hello to you-pl A: hello to you-pl C: I’d like to make an application to [Telecom A: [Sorry! C: to get a free line A: hi C: how long have you been talking on the phone? A: eh, I haven’t been talking for very long C: do you know for how long you’ve been talking?

Here the caller is not satisfied with making an insincere enquiry or with the answerer’s apology (‘‘sorry’’), offer of reconciliation and willingness to listen ("
l ‘come’ ‘‘hi’’) but continues with directly impinging on the addressee’s private space issuing an aggressive request for information (9th turn), which is clearly a reprimand (i.e. ‘you’ve been talking for too long’). Interestingly, the off-record utterances used reinforce the threat to the face rather than mitigating it, as they are standardly expected to do on theoretical grounds. The complaints expressed in (13)–(15) can be explained in GTVH terms as follows: The SO is ‘normal/abnormal’, ‘presence/absence of interest (or desire) of x to contact y’. This can be further specified as ‘availability/non-availability at the moment caller wishes to establish contact’ [examples (12), (14), (15)] and ‘frequency/ infrequency of contact’ [answerer’s second turn in (12) and (13)]. Besides, the SO in (14) and (15) can be further specified as ‘home/public service’. Interestingly, the LM is ‘‘false analogy’’ in both examples. Notice that (12) contains an additional jab line To K o´ l
"; ‘Have you been solving the Cyprus issue?’ which focuses on criticism of lengthy calls justifiable only in case very important issues (of State) are involved. Hence the relevant SO at this point is probably ‘trivial/non-trivial’. What brings together all cases of complaint is expectedly the TA. Crucially, it is not only the addressee who is being targeted but also a certain type of social behaviour, often adhering to the social convention of lengthy chats on the phone. In the examples of the following section the trigger of the complaint is specifiable within the SO KR as ‘success/failure of recognition’. 4.2. Failure of recognition Since overt self-identification is a dispreferred strategy in Greek (see Section 2.1), enquiries like o o& "ı´  ; ‘who is it?’, by giving proof of inability to recognize the caller’s voice, frequently elicit teasing and joking instead of straightforward responses, (providing further clues for recognition). Failure of recognition may reflect on the relationship, and playfulness defuses the seriousness or embarrassment involved. ‘‘Voice recognition can . . . be involved in some of the more creative uses to which people put the communicative affordances of

758

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

telephones’’ and facilitates playing ‘‘intimacy at a distance’’ (Hutchby, 2001: 108). A case in point is the following:

(16)

C: ring A: parakalo
; C: poio& eı´nai; A: pa
nto& o´wi o eo´& C: asteia
kia e; kala
eı´sai; C: ring A: please? C: who is [it]? A: at any rate, it’s not God. C: little jokes, eh? Are you well?

The answerer’s second turn playfully asserts that his identity is to be detected in the caller’s immediate environment, indicating confidence that this second chance should suffice as a clue for recognition, which does in fact prove to be the case. In the following exchange (17), it is the answerer who fails to identify the caller [certainly a commoner case than (16)]:

(17)

C: ring A: nai; C: e
la re; A: poio&; C: kala
ou
te tZ fonZ
mou den anagnorı´zei& pia; A: poio& eı´nai, re ma
stora; C: o Ko
sta&, re anaı´syZte. C: ring A: yes? C: come on re.part A: who [is it]? C: well, you don’t even recognise my voice any more? A: who is it re.part governor (mate)? C: Costas, re.part, you insensitive [ass].

Failure of recognition is first handled with a mild complaint, possibly for prolonged absence of contact (  ‘any more’), to which the answerer responds with an in-group term of address (" 
o ‘re.part governor’) so as to partly fulfil the addressee’s expectations, though still requesting his actual identity. The caller is now

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

759

forced to self-identify and proceeds with a straightforwardly insulting term of address (‘‘you insensitive ass’’). In example (18), the caller responds playfully, as in a game, instead of self-identifying, confident that she is speaking to the intended addressee.

(18)

C: ring A: nai; C: nai, ZmZ
trZ esu
; A: nai, poio& eı´nai; C: den kata
labe&, e; kala
, ta sZmeio
no ego
auta
. A: e
la Ioa
nna, se kata
laba, pla
ka sou ka
no. C: a, eı´pa ki ego
. Ti ka
nei&; C: ring A: yes? C: yes, Dimitri is that you? A: yes, who is [it]? C: you haven’t understood, eh? Well, I make a note of these [things]. A: come on Ioanna, I recognized you. I’m just pulling your leg. C: and I was wondering. How are you?

The humorous utterance appears in the caller’s second turn " 
l"&, "; l´,  " !
! "!
a
‘you haven’t understood, eh? Well, I make a note of these [things]’. The SO should contain something like ‘trivial/non-trivial’, pointing again to overstatement and exaggeration [see discussion of example (13)]. The caller’s pseudo-threat to impose sanctions on the addressee makes him state that failure of recognition was only assumed (l
 o 
! ‘I’m just pulling your leg’) and that it was, therefore, a straight contribution to the game. To appreciate the next example (19), one should bear in mind that the caller is not simply repeating the answerer’s o o& "ı´  ; ‘who is it?’ but also imitating her tone of voice:

(19) C: A: C: A: C: A: C:

ring nai; Xristı´na esu
; nai, poio& eı´nai; poio& eı´nai; o a
no& eı´nai. e
la moro´ mou de se gno
risa. e, be
baia po
& na ma& gnorı´sei&;

C: ring A: yes?

760

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

C: A: C: A: C:

Christina, [is that] you? yes, who is [it]? who is [it]? Thanos [it] is. come on my baby I didn’t recognise you. eh, of course how could you recognise us? (no wonder you didn’t recognise us)

In the same turn, the caller also identifies himself using first name accompanied, however, by the 3rd person of ‘‘be’’ ("ı´  ‘is’) echoing the construction the speaker produced (instead of NP+1st person singular). To the answerer’s admission of failure of recognition, the caller produces a complaint using the first plural to imply distance [as already explained in connection with (12)]. If the Greek telephone game involves two players whose main purpose is to socialise, then a major goal for them is to reinforce links with each other despite lack of visual contact (see Mey, 1993: 198–199). This obstacle constitutes, in fact, one of the main triggers of joking, as speakers feel they should be recognised immediately upon offering a minimal voice sample. ‘‘People play recognition games over the phone’’ (Sacks, [1972] 1995: 550). If the co-player fails to recognise the other on this basis, s/he has scored lower than the game allows and is, therefore, to be ‘‘penalised’’ by the speaker who has made the first move. This is often achieved through teasing, to which the ‘‘victim’’ is expected to respond in an equally jocular manner, so as to finally recover lost ground through successful recognition of identity plus laughter to show appreciation of the ‘‘penalising’’ comment. For this to be the case, however, a presupposition should be satisfied to the effect that the interlocutors are actually close enough to be legitimate players. If either of the prospective players does not have ‘‘the requisite thoughts and feelings’’, any move (or attempt) at initiating the game may abort. Perhaps due to the particular sociocultural background of our subjects (typical Greek students), the recognition part of the game may be so extensive as to become a self-contained unit within the overall structure. So, devoting even a dozen turns to this unit is not unusual. ‘‘Hide-and-seek’’ is also evidenced, as in a case where eight turns precede the transactional part (if any) and the caller pretends to be making an obscene telephone call. He starts his contribution with sighs, moans and the like, only to be recognized after his second turn by the answerer who calls him to task M

o, "
l o´o ‘Mano, come on, cut the crap’. Two more turns follow with the caller asking how he was identified and finally expressing his disappointment with ‘‘fuck you [sic] 
"l  o 
! 
 ’’ ‘fuck you, I just wanted to play a trick on you’. 4.3. Switchboard requests Playfulness is also evidenced in switchboard requests, where it is least expected. In such cases it is commonly initiated by the answerer, who is clearly not the intended addressee.

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

(20)

761

C: ring A: nai; C: tZn 0Anna ya Z
yela A: po
& tZ ye
lete akribo
&; cZtZ
Z
tZganitZ
; C: po
& eı´ste kurı´a Ele
nZ; C: ring A: yes? C: I would like [to talk to] Anna A: how would you like her exactly? Grilled or fried? C: how are you Mrs Eleni?

The answerer in this case transfers an appropriate switchboard request to a restaurant context, where the same expression would refer to how a customer would wish their fish cooked.22 It may be the case that the answerer is implicitly criticising the caller for using  
"l ‘I would like’, instead of some other alternative.23 On the other hand, it may reflect the answerer’s annoyance at being treated as a mere intermediary, not being recognised or even greeted. Thus an appropriate request is responded to with an infelicitous enquiry, as also evidenced in (21).

(21) C: A: C: A: C: A: C: A: C: A: C: A:

ring nai; tZ Marı´a ye
lo tZ ye
lei& polu
; o´wi polu
to´te de stZ dı´noume, tZ dı´noume mo´no se o´sou& tZ ye
loun polu
. ring yes? I want [I’d like to talk to] Maria do you want her very much? not very much then we won’t give her to you [pass her on], we give her only to those who want her very much.

22 The meanings of e
l! ‘want’ are: 1. feel, express a tendency for something, 2. express a strong desire, seek, 3. express a preference, an option and 4. feel, express sexual desire or consent to someone else’s expressed desire. Characteristically, among the examples provided for each ‘sense’, only the ones for the last category involve human direct objects. Therefore, the subcategorisational frame itself can be understood as triggering the specific sexual connotation (Leksiko tis Kinis Neoellinikis, p. 585). 23 Other options in this context would be o ı´ "" lit. ‘will you give me’,  oo
/ 
"l   l
! . . . ‘I would like/Could I talk to . . .’ (see Bakakou-Orfanou, 1988–1989).

762

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

The answerer here probably plays with the sexual connotation of "
l! ‘want’ and playfully refuses to perform as requested. In neither (20) nor (21) are answerers treated as ratified participants, but because of their powerful positions they playfully exploit their control of access to the intended addressee (see Sacks, [1972] 1995: 547). These last examples are reminiscent of the playfulness and concomitant length exhibited in playful reactions to summons responses [section 3, example (19)]. Our proposal, once again, is to interpret them in terms of Giora’s (1997: 185) GSH, since the meaning of "
l! ‘want’ selected in the humorous utterances is evidently considerably less salient in the given context than the one used in switchboard requests. In terms of SO, (20) can be interpreted as ‘actual/non-actual’, ‘switchboard/restaurant service request’ and (21) as ‘expressing desire to speak to x/to have sex with x’. Accordingly, the LM KR is probably ‘‘ignoring the obvious’’ (Attardo, 2001: 27). The similarity with examples (5)–(8) can only show in a specification ‘pun’ under the LA KR. However, unlike in almost all other cases discussed here, no attack on conventions can be detected in (21), therefore, the TA KR will probably have to contain only ‘addressee’. The humorous utterance in (20) is crucially issued by an adult, therefore, it may well be interpreted as targeting the addressee’s failure to abide by conventional means for a switchboard request. 4.4. Formality markers Before concluding, we will consider very briefly the playful exploitation of formality markers, as it is very common practice in this data. In examples (12) and (19), the interlocutor who expresses a complaint either for prolonged lack of contact (12) or failure of recognition (19) uses the plural of the personal pronoun to encode distance, which is actually the trigger of the complaint. In fact, interlocutors use an atypical 1st person plural for self-reference rather than the formal 2nd person plural for single addressees. The implication is that their addressees have distanced themselves to such an extent that formality would be more suitable. In a different exchange, the caller’s first verbal turn is H ı´ Poo
lo Eı´; ‘Is this Mrs Papadopoulou Eftichia?’. Here not only is the answerer addressed with both surname and first name, but also in this order, as if the context were a very formal exchange in the army or an old-fashioned civil service encounter. To this, the answerer responds in an equally formal manner  ı´ ‘speaking’ (formal, old-fashioned, lit. ‘‘the same’’), apparently having failed to recognise the caller, who now proceeds with his second turn using the formal plural D" "  !ı´ " o´; ‘you haven’t recognized-pl me yet?’. A couple more turns follow in the same tenor and the joke actually comes through when the caller states D" "
"" l
 ´  ı´ "& ı´ " "
ı´ ‘your-pl memory isn’t very good because we were having coffee together yesterday’. Here, the formal plural is directly juxtaposed to (and therefore clashes with) the clue offered for recognition, implying an informal relationship between the interlocutors. Only then does the exchange resume its expected tenor: 0El G !
o. D" "ı´"   " 
" & o´ o 
o ‘Oh, hi George, I didn’t expect you-sing to ring so soon’. Similarly, in the following example, the caller identifies himself with D" "
& Po´olo& I!
&, 3o

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

763

´ o& " o
, l
!   
. . . ‘Corporal Papadopoulos Ioannis, 3rd Infantry Battalion, I have the honour to . . .’ as if he were at his regiment in a formal encounter with a higher ranking officer, while he is in fact calling his cousin, who responds in an equally jocular manner and later congratulates him jokingly for being a member of the ‘‘glorious’’ Greek army. The tenor may also change in the opposite direction through the use of slang and abusive terms of address such as l"
o (slang for ‘mentally disturbed’) and accompanying particle ("), !lo"o´ ‘lucky ass’, !
o ‘lit. animal/idiot’, o´ ‘stupid’ !
o 
‘you fat cow’, etc. Games played in this manner are reminiscent of the insult games played by black American children (Labov, 1972) and generally common in teenagers’ conversations. They are obvious examples of ‘‘mock impoliteness’’ (Culpeper, 1996) since they are reciprocated and are neither intended nor perceived as offensive. The exploitation of register for humorous purposes seems to be a significant characteristic of Greek discourse, with the notable help of the old diglossia situation (see e.g. Canakis, 1994). An analysis of register humour in these data is attempted elsewhere (see Antonopoulou and Sifianou, 2002). At this point, it should be simply acknowledged that the GTVH provides the means for such an analysis, as it recognises script oppositions holding between registers or register and subject matter.

5. Discussion and concluding remarks In this paper, we have presented a proposal as to how playful exchanges in telephone call openings could be analysed. Although the data is drawn from a specific institutional context and the subjects belong to a specific sociocultural group, they seem to us fairly representative of dialogic material to legitimise at least tentative conclusions. We have had the double benefit of considering a type of data exhaustively analysed by Conversation Analysis and of applying a very promising theory of humour. We have experimented with a type of material this theory was not originally designed to cater for, but for which it can prove very illuminating, as we have hopefully shown. The points of general interest which have emerged can be summarised as follows: 1. Despite the institutional nature and organisational rigidity of telephone conversation openings, they seem to indicate that conversational material is not always segmentable into distinct sequences of adjacency pairs. As Mey (1993: 252) remarks, ‘‘once we take a look at real conversation, we see that the paired sequences (greetings, questions-answers, requests-compliances, etc.) do not account for anything like the majority of interchanges’’. For instance, in examples (2)–(9), the third turn comes as a reaction to the second one, constituting in this sense an adjacency pair. However, this second turn is traditionally regarded as forming a pair with the summons. In fact, this seems typical of conversational data where interlocutors co-construct the end product through alternation in exchanging turns (see Mey, 1993: 9). In this sense, dialogues are in sharp contrast to story

764

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

telling, where the development and the end of a story is typically under the control of one person, that is, the narrator. If identification of adjacency pairs presents problems, even in institutionalised discourse such as telephone conversation openings, it is evident that segmentation of longer exchanges in less institutionalised contexts becomes exceedingly difficult. Consequently, the distinction between jab and punch lines seems to require further elaboration of the criteria proposed and possibly addition of others, at least for the purposes of humorous dialogue. For instance, in examples (5)–(8) and (20)–(21), the speaker pretends to misunderstand and recontextualises the preceding utterance so that it presents a script opposition with its actual context (Norrick, 1993: 21). The recontextualisation is triggered by the inherent semantic vagueness or ambiguity of a specific lexeme (l!
‘please’ and "
l! ‘want’, respectively), as is also the case in punning. It also transpires from these examples that punning can be seen as representing ‘‘a reaction to a previous turn’’, as Norrick (1993: 65) points out. However, our data do not bear out his other remark to the effect that puns occur ‘‘most obviously in the second position of an adjacency pair’’ (Norrick, 1993: 62). In other words, puns appear to be reactions to something that has preceded but which need not be a first turn of an adjacency pair, for the added reason that segmentation into distinct sequences of adjacency pairs is problematic. Moreover, puns have been attributed aggressive nature and disruptive function in that they misconstrue and ‘‘redirect the flow of talk and action’’ (Norrick, 1993: 64). Wordplay in our data (whether punning or not) has been shown to be nondisruptive, since speakers achieve identification/recognition at these points, as predicted by CA. So what could be suggested here is that there are degrees of aggression and degrees of disruption. A clear case of minimal, or even no disruption is example (1), where playful greetings occur in the position allotted to greetings, followed by initial enquiries as expected. As discussed earlier, the level of disruption is also low in the examples including wordplay on the response to summons. Similarly, failure of recognition can also be low in disruption [e.g.: (16)]. Longer exchanges such as (17)–(21) can be seen as ranking higher on the aggression and the disruption scales. Interestingly, examples like (12)–(13), although high on the aggression scale, do not rank as high in terms of disruption, since CA allows for contact related preemptive moves (Schegloff 1986: 138). Therefore, it seems reasonable to posit only high ranking disruptions as places marking natural segmentation of dialogue. 2. Contextual information and paralinguistic features appear to be fundamental for the construction of the relevant scripts. In our data, the construction of the first script depends on what is conventionally expected in the specific institutional setting, while what is activated by lexical items contributes to the construction of the (partially overlapping but opposed) second script. Situational scripts, that is scripts depending on purely contextual and not necessarily lexical information, are available in the GTVH and used for analysing register humour. In our data, for instance, the ringing of the telephone can be seen as activating the script ‘telephone interaction’. Imitating a person’s tone of voice, as in (19), or intonation pattern employed, as in (13), also contribute significantly to the meaning of the utterances

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

765

discussed. To paraphrase Mey (1993: 241), any utterance, even the most unexpected ones, becomes highly meaningful when placed in the appropriate context. More empirical work is probably required for the GTVH to accommodate conversational data, as it seems that, at least in principle, the theory is compatible with such analyses. 3. In the case of telephone call openings, CA has proved to be very useful for the specification of the content of certain situational scripts, as discussed above, in addition to the specification of KRs such as NS and SI. The SI is constructed with the aid of the conventions detailed in CA, which also constitute the basis for a partial specification of NS, that is, the kind of speech acts performed. The LA KR which seems to us underspecified in the GTVH is also relegated to a secondary position by CA. In view of the amount of playful exchanges in our data and the aggression involved, we are also interested in considering their sociopragmatic function, which lies obviously outside the mechanisms used. In addition, in a society where language is frequently seen as a form of play (Hirschon, 1992; Mackridge, 1992), the function of joking receives further significance. To this end, it seems useful to resort to politeness theory, for which joking serves bonding purposes, an obvious priority for any social group which is positive politeness oriented (Sifianou, 1992). Specifically, on Brown and Levinson’s (1987) account, joking may be used to emphasise shared background and values and also to attenuate face-threatening acts. Interestingly, ‘‘echo’’ reactions to summons-responses [examples (5)–(8)] may well be interpreted as face-threatening acts since they amount to criticism. On the other hand, complaints, which abound in the data discussed, while by definition among the most face-threatening acts, are in fact directed at both aspects (positive and negative) of the addressee’s face. One could suggest here that the face of the speaker is likewise threatened. Thus, the speaker who has a negative evaluation for the addressee has to consider carefully if and how to express his/her frustration and annoyance, because the situation is very delicate. Not expressing these negative feelings would be the most polite way of behaving: ‘‘don’t do the FTA’’ in Brown and Levinson’s (1987: 60) terms. However, such behaviour will not correct the situation and restore balance in accordance with the injunction. On the other hand, if the speaker decides to perform the face-threatening act, s/he has to decide between performing the act ‘‘on record’’ with or without redressive action, or ‘‘off record’’ (see Brown and Levinson, 1987; Olshtain and Weinbach, 1993). It is of interest that our subjects prefer off-record utterances, but of a specific kind; which raises questions for the theory of politeness (cf Kotthoff, 1996: 307). Off record utterances are advantageous for both interlocutors (see Brown and Levinson, 1987; Sifianou, 1997). Since they have more than one plausible interpretation, they enable the speaker to avoid responsibility for having committed a particular act, if s/he considers this necessary. At the same time, they also offer the option to the addressee to attach to the utterance that particular interpretation s/he considers most advantageous for him/herself. One could interpret both ‘‘echo’’ reactions to summons responses (Section 3) and complaints about failure of recognition or contact (Section

766

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

4.2) as off-record utterances. Examples such as (5), (8) or (10) constitute implicit, off-record criticisms of interlocutors’ behaviour, either in terms of convention challenge or genuine readiness to communicate. As in the case of serious key off-record utterances, the addressee has a choice of interpreting the illocution as either threatening or not. In the data under consideration, the choice is between either sticking to the ‘‘social corrective’’ intention and/or to the playfulness (‘‘defunctionalization’’), triggered by and triggering intimacy (Attardo, 1994: 323). Similarly, complaints about failure of recognition or contact are expressed in an off-record manner. Examples such as (12), (14) and (15), for instance, are off-record on more than one levels. The enquiry is neither baldly on record ‘‘You’ve been talking for too long’’, nor on record with redressive action ‘‘Why have you been talking for so long?’’ The latter would leave the addressee the option of providing a serious account for his/her behaviour (e.g.: length of the previous call could be due to settling an important matter rather than to chatting). In fact, callers opt for off-record alternatives introducing additional parameters such as the identity of the previous addressee or the actual length of the previous interaction, parameters which may not be of any real concern to the caller, who is obviously primarily concerned with the difficulty in getting through. The interesting point is that such off-recordness is anything but mitigating the FTA, as the utterances used actually reinforce it. They impinge on the addressee’s privacy and/or call him/her to task in an aggressive though indirect manner. Nevertheless, they are invariably interpreted as playful in the specific context and receive either no response or equally playful reactions. The aggression expressed is part of the game. As Hopper (1992: 185) suggests, drawing on Bateson, ‘‘play emerges in imitation of combat’’.24 In fact, it only imitates real aggression with interlocutors acting roles in the clear understanding that this facethreatening behaviour, far from affecting their relationship negatively, will, on the contrary, strengthen it (see Kotthoff, 1996: 320). This is evidenced in that such interactions unexceptionally resume the ‘howareyous’, frequently produced in ways expressing genuine concern. In other words, in this context, the injunction ‘‘don’t do the FTA’’ has no place. Similarly, negative politeness strategies are not expected and rarely found—quite the contrary, they are occasionally ridiculed. Rather, it is positive politeness attitudes that are conveyed through off-record devices, formed in ways which at first sight seem to be on record. In relation to this context, nobody could suggest that interlocutors choose off-recordness because of the payoffs it affords in minimising face threat. Humour is not used by interlocutors who want to perform serious threatening acts and need to be able to distance themselves from their acts (see Eggins and Slade, 1997: 166). The ‘‘decommitment’’ function (Attardo, 1993: 554–555) is a possibility not exploited, as the threat to be avoided is distance. Humour is not used here by either subordinates to challenge the power structure or by those in power to exercise it subtly (see Holmes, 2000: 178), but by equals who know each other well and use humour as a solidarity-building device. Aggression can be seen as an attack on negative politeness conventions, on formality and distance, as well as on telephone conversation 24

See also Tannen (1990) on the ‘‘ritual combat’’ strategy.

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

767

conventions in general. Jocular abuse is in fact exploited to extremes that verge on rudeness, with frequent exchanges of abusive terms of address; yet no feelings appear to be hurt. Reciprocity in the use of such terms goes hand in hand with the exchange of superficial face-threatening illocutions of all kinds, reminiscent of a table-tennis match. ‘‘Paradoxically, flying in the face of friendly politeness can build rapport, because it signals a relationship which eschews such superficial conventions’’ (Norrick, 1993: 73; Kotthoff, 1996, for an alternative explanation). The ingroup seems to be so tight that the individuals involved have a shared face and can attack and retaliate without the fear of loss of face. Because of this, interlocutors can transform a primarily transactional/referential speech event into an interactional/ social expressive one in order to strengthen their bonds and attack rule-governed behaviour.

Acknowledgements We would like to express our gratitude to S. Marmaridou and to the anonymous reviewer for insightful comments. This paper is part of a longer project funded by the Univeristy of Athens (Special Research Account 70/4/3555 and 70/4/5754).

References Antonopoulou, Eleni, Sifianou, Maria, 2002. To tZlefoniko´ paiwnı´di sta ellZnika
‘The telephone game in Greek’. In: Clairis Chr. (Ed.), Recherches en Linguistique Grecque I: Actes du 5e Colloque International de Linguistique Grecque, Sorbonne, 13–15 Septembre 2001. Paris: L’Harmattan, pp. 59– 62. Attardo, Salvatore, 1993. Violation of conversational maxims and cooperation: the case of jokes. Journal of Pragmatics 19 (1), 537–558. Attardo, Salvatore, 1994. Linguistic Theories of Humor. Mouton de Gruyter, Berlin. Attardo, Salvatore, 2001. Humorous Texts: A Semantic and Pragmatic Analysis. Mouton de Gruyter, Berlin. Attardo, Salvatore, Raskin, Victor, 1991. Script theory revis(it)ed: joke similarity and joke representation model. Humor 4 (3/4), 293–347. Bakakou-Orfanou, Ekaterini, 1988–1989. TZlefonikZ
epikoinonı´a: EkfonZmatikZ
poikilı´a tZ& para
klZsZ& gia su
ndesZ me to kalou
meno pro´sopo ‘Telephone communication: Variation in switchboard requests’. Glossologia 7–8, 33–50. Brown, Penelope, Levinson, Stephen, 1987. Politeness: Some Universals in Language Usage. Cambridge University Press, Cambridge. Canakis, Costas, 1994. Diglossia as an agent of humor in the writings of Elena Akrita. Journal of Modern Greek Studies 12 (2), 221–237. Culpeper, Jonathan, 1996. Towards an anatomy of impoliteness. Journal of Pragmatics 25, 349–367. Eggins, Suzanne, Slade, Diana, 1997. Analysing Casual Conversation. Cassell, London and Washington. Giora, Rachel, 1997. Understanding figurative and literal language: the graded salience hypothesis. Cognitive Linguistics 7, 183–206. Hirschon, Rene´e, 1992. Greek adults’ verbal play, or, how to train for caution. Journal of Modern Greek Studies 10, 35–56. Holmes, Janet, 2000. Politeness, power and provocation: how humor functions in the workplace. Discourse Studies 2 (2), 159–185.

768

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

Hopper, Robert, 1992. Telephone Conversation. Indiana University Press, Bloomington, Indianapolis. Hutchby, Ian, 2001. Conversation and Technology. Polity Press, Oxford. Kotthoff, Helga, 1996. Impoliteness and conversational joking: on relational politics. Folia Linguistica 30 (3–4), 299–325. Kotthoff, Helga, 1999. Coherent keying in conversational humour: Contextualising joint fictionalisation. In: Bublitz, W., Lenk, U., Ventola, E. (Eds.), Coherence in Spoken and Written Discourse. John Benjamins, Amsterdam/Philadelphia, pp. 125–150. Labov, William, 1972. Rules for ritual insults. In: Sudnow, D. (Ed.), Studies in Social Interaction. Free Press, New York, pp. 120–164. Leksiko tis Kinis Neoellinikis (‘Dictionary of Modern Greek’). Manolis Triantafilides Foundation, Thessaloniki. Mackridge, Peter, 1992. Games of power and solidarity—commentary. Journal of Modern Greek Studies 10, 111–120. Makri-Tsilipakou, Marianthi, 2001. Congratulations and bravo! In: Bayraktarogˇlu, A., Sifianou, M. (Eds.), Linguistic Politeness across Boundaries: The Case of Greece and Turkey. John Benjamins, Amsterdam/Philadelphia, pp. 136–176. Mey, Jacob L., 1993. Pragmatics: An Introduction. Blackwell, Oxford. Norrick, Neal R., 1993. Conversational Joking. Indiana University Press, Bloomington Indianapolis. Olshtain, Elite, Weinbach, Liora, 1993. Interlanguage features of the speech act of complaining. In: Kasper, G., Blum-Kulka, S. (Eds.), Interlanguage Pragmatics. Oxford University Press, New York and Oxford, pp. 108–122. Pavlidou, Theodossia, 1994. Contrasting German-Greek politeness and the consequences. Journal of Pragmatics 21, 487–511. Pavlidou, Theodossia, 1995. atikZ
(epi)koinonı´a kai fatika
stoiweı´a. ‘Phatic communi(cati)on and phatic elements’. In: Studies in Greek Linguistics: Proceedings of the 15th Annual Meeting of the Department of Linguistics, Faculty of Philosophy, Aristotle University of Thessaloniki, pp. 710–721. Raskin, Victor, 1985. Semantic Mechanisms of Humor. D. Reidel, Dordrecht/Boston/Lancaster. Sacks, Harvey, 1995. In: Jefferson, G. (Ed.), Lectures on Conversation, Vols. I and II. Blackwell, Oxford (1963, 1970, 1972) (reprint). Schegloff, Emanuel A., 1972. Sequencing in conversational openings. In: Gumperz, J.J., Hymes, D. (Eds.), Directions in Sociolinguistics: The Ethnography of Communication. Holt, Rinehart and Winston, New York, pp. 346–380. Schegloff, Emanuel A., 1979. Identification and recognition in telephone conversation openings. In: Psathas, G. (Ed.), Everyday Language: Studies in Ethnomethodology. Irvington Publishers, New York, pp. 23–78. Schegloff, Emanuel A., 1986. The routine as achievement. Human Studies 9, 111–151. Sifianou, Maria, 1989. On the telephone again! Differences in telephone behaviour: England versus Greece. Language in Society 18, 527–544. Sifianou, Maria, 1992. Politeness Phenomena in England and Greece: A Cross-cultural Perspective. Clarendon Press, Oxford. Sifianou, Maria, 1997. Politeness and off-record indirectness. International Journal of the Sociology of Language 126, 163–179. Sifianou, Maria, 2002. On the telephone again! Telephone conversation openings in Greek. In: Luke, K.K., Pavlidou, Th.-S. (Eds.), Telephone Calls: Unity and Diversity of Conversational Structure across Languages and Cultures. John Benjamins, Amsterdam/Philadelphia, pp. 49–85. Tannen, Deborah, 1990. You Just Don’t Understand. Morrow, New York. Tannen, Deborah, Kakava, Christina, 1992. Power and solidarity in Modern Greek conversation: disagreeing to agree. Journal of Modern Greek Studies 10, 11–31. Eleni Antonopoulou is Associate Professor in Linguistics in the Faculty of English Studies, University of Athens. She studied in Greece (BA Classics) and England (MA and PhD Linguistics). She has published the books Agent-defocusing Mechanisms in Spoken English—A Cognitive Explanation of Impersonalization (Athens, Parousia Monographs 16, 1991) and From the Philosophy of Language to the

E. Antonopoulou, M. Sifianou / Journal of Pragmatics 35 (2003) 741–769

769

Philosophy of Linguistics (Athens, Parousia Monographs 42, 1997) as well as a number of articles in books and journals. Her main research interests are in the areas of Semantics, Pragmatics, Cognitive Linguistics, Humour and Translation. Maria Sifianou is Professor in Linguistics in the Faculty of English Studies, University of Athens. She studied in Greece (BA in English) and England (MA, PhD in Linguistics). Her publications include the books Politeness Phenomena in England and Greece (Oxford University Press, 1992), Discourse Analysis (Leader Books, 2001) and a number of articles in books and journals. She has co-edited Themes in Greek Linguistics (Benjamins, 1994), Anatomies of Silence (Athens, Parousia Monograph Series, 1999) Linguistic Politeness across Boundaries: The Case of Greek and Turkish (Benjamins, 2001). Her main research interests include politeness phenomena and discourse/conversational analysis in an intercultural perspective.