Article body

1. Script-based theory for humour translation in subtitling

According to script-based theory of humour developed by Raskin and Attardo,[1] the mechanism of humour production involves conflicting knowledge representations. A ‘script’ is such an organised chunk of information about something, a cognitive structure internalised by the speaker which provides him/her with information on how the world is organised, including how one acts in it; in the broadest sense it is an object (real or imaginary), an event, an action, a quality etc. (Raskin 1983:199, quoted in Attardo 1994:198 and Attardo 2002:181).[2] Jokes – the main object of study for Attardo – are based on script opposition/incongruity, the use of words which trigger disparate readings, as they are associated with one or more scripts, or packaging of information (e.g. non-sexual or sexual reading of a joke according to the interpretation the punchline forces). The scope of the theory has been broadened by Attardo. Each joke is a six-tuple, it involves the following parameters, or ‘knowledge resources’: language, situation, narrative strategy, target, logical mechanism and script opposition (Attardo and Raskin 1991: 297 and Attardo 2002:176).

‘Language’ refers to choices on the phonetic, phonological, morphophonemic, morphological, lexical, syntactic, semantic and pragmatic levels, which determine the entire makeup of the joke (Attardo and Raskin 1991:297-299). ‘Narrative strategy’ is the micro-genre of the joke[3] (Attardo and Raskin 1991:300). ‘Target’ entails the individuals, groups or the parties in general (that includes ideological targets, institutions that do not have a clear constituency) which are in some way attacked by the humorist (Attardo and Raskin 1991:301 and Attardo 2002:178). As regards ‘situation,’ the joke necessarily has to be about something and the situation can entail the objects, the participants, the instruments, the activities and so on, which constitute the props of the joke (Attardo 2001:5 and 2002:179). Jokes and humorous texts in general exhibit the use of a unique mode of thinking within the universe they create and where rules and logic are defied and applied in bizarre ways. ‘Logical mechanism’ constitutes the resolution of the incongruity present in the joke (Attardo and Raskin 1991:303 and Attardo 1997:409-415). Figure/ground reversal, juxtaposition, false analogies are pertinent examples (Attardo and Raskin 1991:304-306). The last parameter is ‘script opposition’ as was mentioned above. Attardo’s theoretical framework is called General Theory of Verbal Humour (or GTVH for short) and the different knowledge resources are hierarchically organized:

The hierarchical organisation of the knowledge resources (adapted from Attardo 1994:227)

The hierarchical organisation of the knowledge resources (adapted from Attardo 1994:227)

-> See the list of figures

Still, as Attardo has stressed, the production of a joke can be triggered by any knowledge resource, with the rest of them being filled in and the levels presented here ‘do not correspond to the consecutive stages of actual production’ (Attardo and Raskin 1991:327).[4] The model was developed as a formal tool of establishing how different or how similar two jokes can be intralingually; the higher up a difference is traced, the more dissimilar they will be (no value judgement is passed when ‘higher’ and ‘lower’ are used; the terms refer to the model’s linearity) (Attardo 1994:228). This premise was taken a step further and applied, mutatis mutandis to interlingual equivalence, that is, translation of jokes (see Attardo 2002:184-192). A translated joke is an ideal translation when it shares the same script opposition (say sexual vs. non-sexual readings), the same logical mechanism (say analogy or role exchanges), the same situation, the same target (say blondes) and the same narrative strategy as the original (language has to be necessarily changed when translating).

It is worth noting that the definition of script was later partly altered by Attardo, or at least the social aspect of the notion was highlighted:

[scripts are]… collections of semantic information pertaining to a given subject… [embodying] the sum total of the cultural knowledge of a society, which can be represented as a set of expectations and/or weighted choices.

Attardo 1997:402

The social grounding of humour can become even more prominent if the notion of superiority and its relation to incongruity (Attardo’s script opposition) is also examined (Attardo incorporates superiority in his model under the ‘target’ parameter).[5] As Vandaele argues, it is impossible to offer a satisfactory explanation of the field or humour on the basis of one of the two main principles alone (incongruity or superiority) (1996:242). Superiority relates to incongruity in various ways: a) most acts of incongruity can be assigned to a social product and/or agent, which are thus seen as inferior; b) ironic incongruity is controlled abnormality as a sign of superiority; c) incongruity can in most cases be resolved and overcome, thus creating superiority (each time we laugh at humour, we demonstrate our wit to our peers and diminish the social pressure they may exercise on us); d) some incongruities are conventionalized as humorous (conventionally forced via cues, or humorous stereotypes that are supposed to be funny per se) (Vandaele 2002:157). In this broader sense, (of socialization, stereotyping, problem solving and so on) superiority can be diffusely scattered over all six parameters that Attardo suggests (Vandaele 2002:158).

These considerations take us a bit further from a strictly cognitive approach. I will present a theoretical model of subtitling humour which will take on board both Attardo’s knowledge resources and Vandaele’s concerns. I will thus suggest that verbal humour involves social/cognitive expectations, that is, a sort of norm acceptance and/or norm opposition. Norm acceptance is, when, for instance, a stereotype, a cliché, something societies have established as inherently funny is used, (scatological references, national stereotypes etc.). ‘Norm acceptance’ refers to contextual/social factors generating humour and their moment-to-moment assessment and shows that something can be humorous without exclusively involving a clash or incongruity. Verbal humour can – usually simultaneously – involve norm opposition. ‘Norm opposition’ subsumes script opposition, but ‘norm’ highlights the social rootedness of humour. It can involve two clashing interpretations created by a pun, for instance, or the play with tabooed issues (the repertoire of such issues is norm acceptance in its own right) in situations where it is not appropriate (hence the clash). Norm opposition in that respect counts for cognitive, but also for social incongruities (what clashes with what and in what situations can be seen as a social convention). It can equally involve deviations from the ‘natural’ and ‘proper’ use of language – for example, stuttering or taking everything literally, which also creates a sense of superiority for the ‘bystanders,’ that is, the viewers. The screenplay writer can indicate how something which is peripheral, that is, the unconscious use of language/interaction rules can become the focus by flouting them (when, for example, a dead metaphor is revived, or when politeness rules are not upheld). Norm acceptance and norm opposition are two sides of the same coin and they can be structured on the levels (knowledge resources) that Attardo proposes with incongruity and superiority, being diffusely distributed over them. Norm acceptance/opposition in films can be viewed as a means that establishes humorous communication between the director/screenplay writer and the viewers. It is the vehicle to highlight/establish cleverness (sometimes to the detriment of a targeted individual/group/institution/idea), natural understanding, levity, in-groupness (shared experiences and knowledge) and the assertion of a common metalanguage or a shared code of some sort. This does not necessarily mean that within the film-as-such-level characters cannot communicate among themselves in a similar way. Still, even in that case, the viewers can ‘communicate with’ some of the characters and indirectly with the director/screenplay writer who talks over the head of film characters. Communication on this director-viewer level is very prominent in the films in question; in fact, the character’s semblance of seriousness increases the funniness of what is said and renders characters absurd caricatures (targets) who unconsciously fail to follow the rules of social propriety and politeness as well as natural and coherent turn-taking.

In light of all this, and bearing in mind the contextual variables of subtitling, a humour theory model of norm opposition/norm acceptance for this type of translation can be graphically represented as follows:

Key: LM=Logical Mechanism, SI=situation, TA=Target, NS=Narrative Strategy, LA=Language.

-> See the list of figures

The whole construct is characterised by a great degree of circularity, as will be shown. The internal structure of a humorous sequence entails norm opposition or norm acceptance, an abstract ‘social slot’ for something, the intention to use a humorous element and the socio-cognitive convention that is the precondition for its humorous function. Norm acceptance/opposition reflects the director’s or screenplay writer’s intention to humorously communicate with the audience and is structured along the knowledge resources Attardo has suggested. The prominence of the knowledge resources varies, depending on the type of the humorous sequence. ‘Target’ will be prominent in the genre of satire, for instance, and in ironic statements in general.[6] The last knowledge resource is language and it is the realisation of norm acceptance/opposition. The levels on which it can be organized range from the phoneme (an awkward pronunciation) and moving along the axis to sentence level and then beyond that, to larger portions of ‘text,’ which in their turn can be contiguous (as in turn-taking sequences) or displaced (as in the case of repetitions of humorous statements which may straddle longer stretches of text). The last category is register and it can involve any of the preceding linguistic units.

Humorous sequences structured along these parameters and realized on various linguistic levels tap into the context. The contextual factors displayed in the figure, here called externalities, are a rough indication of what the context of a film might be. Image accounts for the polymedial nature of films. Certain actions, objects or entities are present on the screen and humour perception is linked to the plot unfurling on the screen. Constraints refers to what a language can do. Certain words or other textual material are more amenable to jocular use in some languages than in others (lexical, syntactic ambiguity, spoonerisms and so on). Cultural conventions may also dictate which humour routines and stylistic manipulations are acceptable and all that within the spatio-temporal constraints specific to subtitling (lengthy compensation techniques are therefore rare; see, for example, Asimakoulas (2001) for the inconsistency in/insufficiency of lengthy explanations for humorous proper names).

Presupposed knowledge is a vast aspect of context and can embody the encyclopaedic knowledge that people possess individually or collectively, cultural assumptions, or knowledge accumulated by observing/experiencing the world. The term incorporates linguistic presuppositions and non-linguistic ones, both of which are made every time we speak or write; every text is affected by them and contains both linguistic and contextual/cultural triggers (Fawcett 1997:124-125 and 1998:114-118).

Intertextuality is the property of texts depending upon previous instances of texts. Such links between texts can be held between elements of the given text as well as between distinct texts and the generation of intertextual links is usually motivated and relates to the text function/the overall communicative purpose (Hatim and Mason 1990:125,128), which is the humorous effect here. Intertextuality includes allusions, parody and commonly repeated segments which become funny precisely because they recur through the interaction. Humorous intertextuality involves norm opposition in the sense that the ‘grafted text’ clashes with its source, by being different, for example, or by being used in different circumstances. As we will see in the data discussion section, it also involves norm acceptance, the use of a recognisable chunk of information.

The last contextual factor given in the figure is the interpersonal level, which is to do with the expression of a certain attitude and feeling, as in superiority/disparagement humour and satire. The purpose of disparagement humour may not be the venting of one’s anger – hence the interpersonal level here – against a certain group of people. However, it cannot be considered to be completely innocuous either, as Saper observes (it depends on who says what to whom or on whether it is used symmetrically – among members of the same group – or asymmetrically) (1991:233-236). This can again involve norm opposition, if it is unacceptable to make such comments in a given context, or norm acceptance, if it is a recurring humorous device.

In the light of the above, the following mini-theory of humour equivalence can be postulated: things being equal, the ideal aim of the subtitled version of the original dialogue is to reflect as closely as possible the structure of the original humorous sequence, taking into account contextual variables and using the appropriate language. As Widdowson put it, language can be seen as ‘the formal encoding of the most common features of context,’ but it can also ‘project its own contextual implications’ (1998:17,21). The proposed model is an extension of this premise, with the internal structure feeding on the externalities and vice versa. Still, cross-cultural and cross-linguistic differences may dictate shifts of the norm acceptance/opposition scheme on various levels. These shifts can vary from slight changes to total recontextualisation and may involve any linguistic unit. The next section will show how the structure of humorous sequences can vary in subtitles.

2. Data

In this section certain categories of humorous sequences will be discussed. Various translation solutions will be presented, one of each kind where possible (examples are representative of the strategies; the general categories selected for analysis here exhibited roughly the same distribution in both films, possibly reflecting humour consistency in this genre). The original occupies the left column and the Greek subtitled version and its backtranslation is placed in the slot next to it. Slashes indicate subtitle breaks and dashes the absence of a subtitle, while brackets contain metalinguistic (or other) information necessary for the discussion. The source of each sequence is signalled with an NG (Naked Gun) or an A (Airplane!).

2.1 Wordplay

Puns can be generally translated as follows (Delabastita 1993:192-226). 1) Pun rendered as pun: the ST pun is translated by a TL pun; 2) Pun rendered as non-pun: a non-punning phrase which may retain all the initial senses (non-selective non-pun), or a non-punning phrase which renders only one of the pertinent senses (selective non-pun), or diffuse paraphrase or a combination of the above; 3) Pun rendered with another rhetorical device, or punoid (repetition, rhyme, referential vagueness, irony etc.); 4) Pun rendered with zero pun (total omission, or avoidance strategy, so to speak); 5) ST pun copied as TT pun, without being translated; 6) Addition: a compensatory pun is inserted where there was none in the ST (possibly making up for strategy 4 where no other solution was found); 7) Editorial techniques: footnotes, endnotes, comments in translator’s forewords etc.

These techniques apply to subtitling too as the following examples show:

-> See the list of tables

In (1) Jane refers to the prospect of postponing the rendezvous, and Frank absurdly takes it as a concrete counterproposal of some sort, which he declines. The subtitler created a pun by intelligently using a standard expression routinely used for a dinner cancellation acceptance and a commitment-seeking device for a future meeting (‘though’ can be translated in many ways in Greek, but the function here is commitment-seeking). Thus a completely different narrative strategy is used (narrative strategy is used here in the broader, rhetorical presentation-stylistic routine meaning). The funniness lies in Frank’s inappropriate interpretation of the expression as a sex-act promise. The norm acceptance/opposition scheme has changed in that a new social contradiction is exploited (dating procedures). In his detached, innocent way, Frank falsely presupposes that Jane seeks sex before she even gets to know him (role reversal as the logical mechanism), not seeing his priorities (food before sex). Thus, Frank remains the target of this sequence, but the situation is radically changed, as a new range of norms and norm contradictions are exploited (social rules of decorum).

Puns are embedded in larger or shorter sequences and a complex example of this is (2). Firstly, a hackneyed frame (‘boy finds…’) is used, but Frank flouts expectations, stretching this frame and producing a ludicrous expansion of trivia and an absurd portrayal of a love story with an unexpected end (hence the funny incongruity). Orange Bowl refers to the stadium where football games between various American colleges are organised and a ‘blimp’ is a mini balloon used for broadcasting images from above. Goodyear refers to the tyre/blimp producing company or, according to Frank’s innocent interpretation, the calendar year. Jane’s turn here flouts the norm of social propriety, as the question is not what one would normally want to know about the accident. All this cannot be replicated into Greek. A range of changes were made on the situation and narrative strategy levels, as well as the logic of the ‘joke’ bringing in different norm acceptance/opposition schemes. The ‘boy finds girl’ frame does not exist in Greek. This part is rendered as ‘the same story,’ (fixed expression for something repetitive and irritating). The almost identical narrative scheme following this part is highly reminiscent of a line by the Greek national poet, Dionysios Solomos, (‘The mother loses the child and the child the mother’) which also has numerous funny permutations (intertextual links that also constitute presupposed knowledge for Greek speakers). The proper name of the stadium is blotted out in order to save some characters and because it is too culturally specific. As concerns the pun, it is rendered with a punoid, a standard Greek wish, which also serves as a temporal reference (‘upon saying this wish,’ or at 12:00 on a New Year’s Eve), followed by a wish which has undergone antonymic lexical substitution (Leppihalme 1996:201). Interestingly, Jane here seems to be asking not a trivial and uncommon question as in the original dialogue. Her question sounds more like a statement on the tragic nature of the accident exactly because it happened on a New Year’s eve (it is typical to stress the role of fate and/or something bad by using temporal routines in Greek, such as holidays and times of the day/greeting routines, a Greek narrative strategy). Frank’s answer is an enhanced expression of the tragic, building on Jane’s routine, but not an innocent statement as in the original (Frank and Jane are not ‘targets’ here, but playful cynics).

Sometimes the solutions offered do not fare that well and are unnatural, as is the case with (3); the pun ‘nightcap’-drink/‘nightcap’-cloth is rendered as drink ‘a glass/glassful,’ which is the revival of a dead metonymy in Greek. The situation is again changed completely, with Frank being the innocent target who declines a drink invitation because of the obviously bad health implications that eating a glass has. Still, the logical reversal here is very marked in the sense that the metonymy is so fixed, that it cannot be manipulated in this way and therefore becomes less funny (as was mentioned there are constraints, or degrees of acceptability as to what can be done with language).

Although wordplay and pun are used interchangeably by Delabastita (1993), he later indicates the blurred line between wordplay and certain other (and wider) contextual or rhetorical devices (it now covers such diverse phenomena as clichés, metaphors, parodic allusions, spoonerisms, idiom-based puns and onomastic puns (1997:14)). In the films discussed here, there are certain instances of wordplay which involve a norm opposition of a more contextual kind. Thus, verbo-pictorial wordplay can be created by exploiting the relation between various meanings of an expression and the pictorial components of the film (the image externality):

-> See the list of tables

These are instances of ‘understanding and experiencing one kind of thing in terms of another’ (Lakoff and Johnson 1980:5). Some are dead metaphors ‘where one is hardly conscious of the image’ or ‘the picture conjured up in the metaphor’ (Newmark 1988:105-106). This missing image can then be concretised in the actual physical image and the result is a dual actualisation (Veisbergs 1997:158), or the revival of the literal meaning. This literal meaning is used to cancel the figurative meaning, also flouting viewers’ expectations on the unconscious and ‘natural’ way of using language (norm opposition).

Subtitle (4) preserves the structure of the original: same situation, equivalent narrative routine for ‘barking orders’ (conveniently elliptical in Greek) and the same logical contradiction between what is said and shown on screen (possibly the people who execute the orders are the targets here). It even involves triple norm activation; the Greek fixed expression additionally signals that the airway is ‘in the spotlight,’ it is the focal element of attention (and of the plot at the moment the utterance is made). The same does not hold for (7), where ‘bingo’ can be used as a success-signalling interjection (narrative strategy) or as a reference to the game, a presupposed bit of knowledge not available to Greek speakers. Its retention is far from an effective translation solution here. Example (6) is a complex instance of metaphor. Frank counters all rational expectations. For a moment, he seems to be using a metaphoric expression to express his admiration for Jane’s pubic hair, only to be handed an actual beaver, while the use of ‘stuffed’ sustains the metaphor and the ambiguity of sexual/a-sexual references. This is norm opposition in two ways: a clash of what is said and what is shown on screen, as well as inappropriate comments in the given context (overt sexual comment and with someone Frank has just met). This was intelligently preserved in the subtitle by making minor changes on the situation parameter (or the ‘props’): a generic term for (coat) fur was used. Although the use of this general word to refer to a beaver is rather marked, it sounds natural in Greek and it even increases the surprise effect and humorousness. It reflects Frank’s idiolectal, humorous way of referring to things using slang expressions. Similarly, subtitle (5) is deftly rendered with an idiomatic verb, which happens to belong to the same semantic field. The situation is slightly changed, as the verb means ‘to give somebody an earful for their mistakes’ and it therefore refers to the intensity of (verbal) criticism they will receive, not trouble in general.

Generally speaking, wordplay was adequately rendered when the subtitler changed the norm scheme, coming up with creative recontextualisations of the humorous sequence (except for some marginal cases, where mistakes were made or the norms were not satisfactorily manipulated). The majority of wordplay instances did not fare that well though; there was considerable loss of linguistic humour when the subtitler tried to retain the original structure.

2.2 To boldly make…bizarre comparisons

Metaphors and similes (and any comparison) can have a humorous effect when the compared parts exhibit less semantic similarity than expected, or when there is an absence of ‘semes’ between the notions compared (Vandaele 1996:249), which is a sort of norm opposition. Another instance of humorous sequence of this type is when sustained metaphors comprise an arbitrary mixture of metaphors or when new but trivial metaphors are used instead of fixed ones or new, ‘revealing metaphors’ (ibid.).

-> See the list of tables

Frank’s monologue in (8) is an original simile (by analogy to original metaphor (Newmark 1988:112)) and it constitutes a new way to perceive things, but a rather bizarre one. Humour here is attributed to a triple norm opposition complex: an unorthodox comparison is made, a standard expression (bubble up/over) is expanded and a trivial (not usual but equally not revealing) situational analogy is invented. Maybe a verbatim translation would seem odd or it would not have the same force (if the representation is cognitively/culturally too distant, as there are certain conventions dictating what is trivial and what is not, or what is funny or simply strange). The solution given here is an exaggeration of Frank’s state of mind with no particular humorous thrust. The logical mechanism of making an analogy is retained. Still the situational prop and means of comparison (carbonate) is more general, turning the simile into a hyperbole (different narrative strategy). The result of this is the loss of the target component of the original humorous sequence too, as Frank sounds more frustrated about the situation at hand and as if he is exaggerating, rather than as someone who consistently fails to use language appropriately. The subtitler could have created a more original simile, or (s)he could have highlighted the details of a trivial comparison more (by inserting a specific Greek or other carbonated drink, for example or by manipulating the routine of an advertisement). The unoriginal trope change[7] here reduces the force of the original.

The same tendency to dilute (unoriginally replace) the figurative expression is observed in (9) where a generalizing translation was employed, that is, the replacement of ‘tylenol’ (popular painkiller in the USA) by a hyperonym (Hervey and Higgins 1992:95). Throughout the film, Jonny consistently contributes a series of irrelevant bits of information whenever he is asked, paying unusual attention to detail. The slightly changed situation variable (pill) reduces the humorous effect, because it is not detailed enough to highlight the clash with the preceding description. Any commonly available/known Greek pill with the same or different connotations or properties (white/elongated) or a hyperonym with different humorous connotations could have increased the humorous effect and be consistent with Jonny’s characterization.

The same ‘dilution’ occurs in (10) where a complex interlocking sustained metaphor (which gradually develops to a multiple norm opposition) is used: ‘gloves’ refer to boxing, ‘hard ball’ to baseball, and ‘full court press’ to basketball. The conveniently shorter subtitle may save a few characters, but exhibits only one unusual juxtaposition of concepts and therefore a shift occurs on the situation level of the sequence. The commonplace metaphor ‘play the game wildly’ (to take decisive action/fight fiercely) is concretised. This remotely reflects Frank’s warning/challenge to Ludwig (the evil machinator of the film) whose name (a vocative, part of the warning) is also blotted out in the subtitle. The second part of the sequence could be perceived as a less funny stream-of-consciousness comment or follow-up of the initial warning and not the narrative strategy of an all-out sport challenge.

In (11) the structure of the humorous sequence and the proper name of the original is retained in the subtitle (Drano is a popular toilet and sink clog remover in USA). Viewers’ presupposed knowledge is not taken into consideration and the emotionally drained/physically clean incongruity is not obvious at all (perhaps a similar Greek product or a drug with cleansing properties could have been used).

For some reason, humorous comparisons were not translated successfully in the two films. The norm scheme was altered, but it involved unoriginal/less humorous comparisons.

2.3 Clinamen… The marriage of allusion and verbal parody

According to Rossen-Knill and Henry, verbal parody is a highly situated, intentional and conventional speech act which represents the object of parody (which can be anything: events, actions, beliefs, thoughts, individuals, groups, institutions and so on) and flaunts that representation in order to criticise that object in a humorous way (1997:721,740). Three levels or parodic scales will be postulated here. On the presentation level, the parodying text can be identical or modified and its modification may involve skewing (replacements) or expansion (additions), quantitatively speaking, and style or content alterations, qualitatively speaking (or both). On the specificity plane, the alluded text may be a specific textual entity or a diffuse one (the source is not readily retrievable, or a pastiche of styles and contents related to a genre/an activity are involved). The third plane is to do with the norm acceptance/norm opposition scheme; norm acceptance could be the recognition of a hackneyed topic or a clichéd style and norm opposition involves, for example, the stylistic incongruity between a parodical sequence and its source text or between the elements of which it consists. This double recognition (form and source, or presentation and specificity) establishes the relation between the screenplay writer and the viewer; just as in the case with writers and readers, parodic allusions incite viewer engagement and co-operation, with the viewer tracing allusions and intertextual references (Leppihalme 1996:202); thus an appreciative reader (here viewer) is included in an in-group of some sort (ibid. 1997:49):

-> See the list of tables

Example (12) is a recognizable trite formula used by policemen in the USA. As Vandaele notes, this stereotypical utterance evokes an interpretation (incident attracting many curious spectators), which is flouted by what we can see on the screen (only a handful appear) (1996:250). The same formula is used later on when a missile-bearing truck crashes into a fireworks store and umpteen explosions occur. Humour is saved by what is shown on screen, but the parodic thrust is lost in the subtitles. The structure of the ‘joke’ here changes, because the subtitles lose the target component (police formulae alluding to police incompetence as a group/institution). The use of a rather formal variant of ‘unusual’ in the second subtitle renders the statement a funny, intentional ironic comment[8] or an exaggerated understatement (a vain attempt to downplay the importance of a ‘watchable’ accident).

In other cases, a loosely evoked text means that the subtitler has more leeway when manipulating the target version, as in (13). The anticlimax of mentioning stoplights enhances norm opposition (they should normally be noticed at all times) and norm acceptance (Frank’s consistent characterization as a caricature). The subtitle describes a slightly different situation, by changing the angle of consciousness: Frank talks incoherently (incoherence is part of the Greek stereotype of romantic raving in popular films), and not about the things of whose existence he was unaware before. The target here is not just Frank and his poor driving skills, but romantic, pseudo-poetic reciting in general. Narrative strategy is still similar (pseudo-poetic style) and there is a deft and abrupt transition from poetic wording (‘newly-sprung’) and an affective diminutive to a blunt, realistic and more explicit description (‘at the junctures’). This explicitation may compensate for the omission of ‘honest’ further down (possibly to save some space) which eliminates the target component (police corruption).

Let us now examine examples of modified and specific parodic allusions, which are more difficult to tackle (original frames underlined):

-> See the list of tables

An expanded and skewed version of a Dirty Harry (1971) dialogue (underlined) in (14) is inaccessible to Greek viewers. Allusion is lost here too and the situational props are unsuccessfully changed, reducing the appeal to viewers’ encyclopaedic knowledge. The replacement of ‘togas’ with stockings was possibly intended to highlight the ‘weirdness’ of the situation, or maybe to evoke a theatre stereotype (Robin Hood costumes perhaps) or a bank robbery situation, but it is confusingly obscure and odd (especially when taking the evanescence of the subtitle on screen into account). It is not clear how stockings match with Shakespeare (in the way that ‘weirdos’ ‘togas’ match with Julius Caesar in the original). The same loss of intertextual thrust applies to (16), which in Casablanca (1942) reads: ‘it doesn’t take much to see that the problems of people don’t amount to a hill of beans in this crazy world.’ This frame is skewed and expanded with the figurative expression being concretised. The subtitler took an equivalent Greek idiom indicating the insignificance of their problems as the point of departure and subsequently deidiomatised it (Veisbergs 1997:159). The double recognition of the alluded frame is lost (norm acceptance/opposition) but a different norm scheme is used describing a different situation but with the same target, Frank, and his way of speaking.

Another bizarre reframing occurs in (15) and for a text with a very high communicative currency in the USA, the national anthem. A tone-deaf and ignorant Frank (disguised as a tenor to find the queen’s assassin), finds himself in the awkward situation of having to sing the national anthem in a full stadium and distorts it.[9] There is no subtitle for this sequence, although Ivarsson argues that songs which have a bearing on the story should be translated at least partially, as an indication of what kind of song it is (1992:119). Thus, the best part in this film (according to viewers’ comments <http://www.imdb.com>) is lost. The reframing of the Greek national anthem could perhaps offend many Greek viewers and was avoided.

Parodical intertextual references in the films were complex and did not survive translation, especially for sequences from specific alluded texts. Non-specifically derived parodic texts proved to be easier to manipulate and the subtitler seemed to have taken liberties and to have ‘enhanced’ the humorous effect, but still certain inconsistencies were to be traced.

2.4 Who’s Afraid of Disparagement?

Disparagement or superiority humour presupposes a norm acceptance, a presupposed cultural convention,[10] or norm opposition, that is, degrees of acceptability when it is used in certain contexts:

-> See the list of tables

In (17) Frank expresses what is typically ineffable in a speech: he unreservedly expresses a negative attitude towards monarchy. Register incongruity is also involved as ‘silly’ contrasts with ‘gracious, considerate hosts.’ The subtitle changes the target, sounding more than a personal attack on the queen herself, possibly the way she looks/dresses, which is not an uncommon comment in Greek newspapers and comedy programmes; still, this norm could be further exploited by adding a short, clear comment on the queen’s appearance, for instance. The linguistic clash mentioned above is not exploited as the subtitle is consistently formal. In (21) ‘Turkish prison’ happens to have an additional humorous effect for the TL audience because of the broad (negative) stereotype repertoire the Greek viewers possess with regards to Turkish politics and institutions.

The rest of the examples exhibit the use of the same situational variables and same narrative strategies, but the target level seems to be a bit more problematic. In (18) the subtitle does not have exactly the same connotations, as a building contractor is a respected professional in Greece who does not fit the stereotype of having a particularly violent behaviour. Replacing this target with ‘builder’ or ‘butcher’ would be more successful. Example (20) possibly refers to the dearth of important Jewish athletes or to their mediocre performance and it is part of the tradition of a rather large repertoire of popular Jewish jokes in the USA. This target-convention is again not readily retrievable for Greek viewers and this aspect of the joke is lost, but the overall effect is retained in the lightness of the leaflet shown on screen, which is a concretisation of the dead metaphor. Finally, in (19) each bilingual display serves the ridiculing of the Spanish community in the USA.[11] In the displays in question this (arbitrary) code mixing and the use of some transparent words (‘putanna’) is very funny, but they are eliminated in the subtitles where the standard Greek language of signs is used.

Generally speaking, the subtitler was unable to successfully convey disparagement humour, except for ‘accidental’ instances of stereotype overlap. By retaining the same targets of the original sequence, the presupposed knowledge of viewers was not taken into account.

2.5 Register-based Humour

The following examples indicate how language varieties can be manipulated to create humorous effects, perhaps one of the intractable problems for subtitlers:

-> See the list of tables

These are instances of sociolect, or ‘language varieties typical of the broad groupings that together constitute the ‘class structure’ of a given society’ (Hervey and Higgins 1992:118). The humorous device employed in (22) and (23) is the subtitling of the original ‘inscrutable’ Jive talk into standard English. As when translating dialects in general: ‘the class structures of different societies, countries and nations never replicate one another…[and] there can be no exact parallels between sociolectal varieties of one language and those of another’ (ibid.:119). In subtitled sequence (22) the first few turns are marked as informal (use of contractions, slang/idiomatic language) and the rest as formal. Thus, one interlocutor adopts an informal style only to be answered to with slightly ‘loftier’ Greek by the other (there is, however, some inconsistency when they both speak informally at the end). Instead of Jive speakers as a group being the target here, the inability of one of the speakers to be on the same formal-informal axis with his interlocutor is targeted. Still, in (23) no alternative solution is offered and equally no explanation is offered why the stewardess is puzzled by the ‘cryptic’ sociolect of Jive (the subtitler could have used the same technique as in 22 and have one of them speak in formal Greek and another in informal).

The distance between formal-informal language is also focused on in subtitle (24). The subtitle describes a slightly different situation than the blow by blow description of the original (omissions are underlined) and does not exhibit the same pastiche of styles as the original. Humour in the subtitles relies on this clash between a solid stretch of detached medical jargon and then an informal, sentimental comment at the end. The victim description at the end signals the emotional climax/personal comment (notice the informal wording and the diminutive use) of a doctor who made a very formal and detailed diagnosis a few seconds ago. In fact, the diagnosis is even more formal in the subtitle; the formal ‘uncontrollable saliva emission’ (underlined in the subtitle) was added, also on the basis of what is shown on screen (the situation’s props are changed here). The host of the virus actually starts to drool and the subtitler seems to have cleverly taken the image factor into account here. This insertion thus increases formality and highlights the humorous development of the symptoms in real time (Dr Rumack remains the target here because he fails to see that the symptoms develop as he speaks). The funny comparison at the end is once more lost.

The target versions generally reflected the attempt to fully exploit opportunities for successful humour translation on the part of the subtitler. Some omissions did occur though, possibly reflecting a general tendency in subtitling: since colloquial expressions and dialects are difficult to render and are usually eliminated and reformulated into standard wording and expression, because irregularities may be perceived as mistakes or may impair comprehension (Smith 1998:145,146).

Conclusion

This paper has attempted to provide the tools for the description of humour translation. If humour is compartmentalised as in the suggested model, it can be described more efficiently, a practice potentially pointing to translation solutions. The subtitler can ask some questions on the type of humour that is involved. For example, is it to do with breaking a social convention such as politeness? What logical mechanism is used (reversal)? Is it about a targeted group/nationality? In what situation? What narrative strategy is employed? How is language used to describe it, on what level and how can I move up and down the levels to achieve the best result and compensate for losses? Which externalities come into play and what is presupposed by viewers?

Breaking humour down into components can be a useful practice and an exercise of finding alternatives. Translating humour requires creativity and in the films in question, changing the structure of the humorous sequence proved to be an effective technique. This applies even to register-based humour where long stretches had to be edited by the subtitler, but also (verbo-pictorial) wordplay involving much smaller units. Parody was generally lost, but an attempt was made to recreate its effect playing with language and register. Humorous comparisons and disparagement humour proved to be the categories that were tackled least effectively.

Subtitles have to be concise and clear, because space and time restrictions are simultaneously at play. If a subtitled humorous sequence is bizarre (as was the case in some of the examples discussed) the viewer does not have the opportunity to process it and enjoy humour after the subtitle is flashed off screen. Still, the inconsistencies, toning down of humour and downright mistakes observed cannot only be attributed to the difficulty of rendering humour in a concise and clear way. Other constraints, such as lack of talent/experience, insufficient training or working conditions (including access or not to reference material, commission deadlines and fees) can also have a bearing on the final product and lead to solutions or non solutions for translation problems.

Concluding then, the translational problems/idiosyncrasies related to humour require further investigation. The various functions of humour need to be looked at in the light of various genres and media constraints. In certain cases regular patterns may be observed as in Jaskanen (1999), where an attempt was made to shed light on the translational norms (norms as used in Translation Studies) operating for subtitled comedy films in Finnish. All this also has to be combined with a more global approach; since humour is where the translator’s creativity is put to the test, the external factors that affect this creativity have to be investigated. Some interesting directions would be a) the investigation of the sense of humour of individual subtitlers; b) the ways in which individual subtitlers’ sense of humour surfaces in their work; c) the conditions under which talent and efficiency can be improved, possibly in training programmes for subtitlers which will address humour translation and broader creativity issues. Whatever the angle adopted, there is great potential in this underresearched area for a wide variety of studies enriched by interdisciplinary cross-fertilisation.