Revista Alicantina de Estudios Ingleses 19(2006): 191-210 Multimodality and Relevance in the Athens 2004 Olympic Games Televised Promotion*

This paper is concerned with a multimodal communicative act, the televised promotion of the 'Athens 2004 Olympic Games' on National Hellenic television. The first aim of the paper is to show that the commercial examined constitutes a "multimodal metaphor" (Forceville, 2004 and 2005) through which the audience is essentially invited to interpret the 'Athens 2004 Olympic Games' in terms of a 'festival'. The second aim of the paper is to explain how it is that a commercial such as this one, which foregrounds entertainment and celebrations, is almost unanimously received by a multicultural audience as successnilly advertising the Athens 2004 Olympic Games. In the framework of Relevance Theory, the pragmatic and cognitive benefits of the specific multimodal metaphor are discussed in terms of positive cognitive effects and processing effort (Sperber and Wilson, 1986/1995; Wilson and Sperber, 2004). Experimental evidence on native and non-native audiences' reaction to this televised promo suggests that the intended interpretation is recovered in terms of a range ofstrongly implicated assumptions that viewers unanimously reckon, rather than in terms of a range of weak implicatures (Forceville, 1996). Drawing onthe Cognitive and Communicative Principies of Relevance and the key concept of Optimal relevance, the paradoxical finding of 'irrelevant scenes' contributing to a fruitful interpretation process and recovery of 'relevant meaning(s)' is explained.


Introduction
Upon watching the televised promotion for the Athens 2004 Olympic Games, viewers are likely to feel a spirit of joy, exhilaration and enthusiasm forcefully conveyed by the commercials promoting the particular product.Intuitively, we would expect that viewers irrespective of socioethnic origin are also likely to get carried away by these commercials that invite them to see sports as fun, and the Athens 2004 Olympic Games as an opportunity for entertainment and celebration.
This paper focuses on one of the commercials of the above campaign with a twofold aim: firstly, to argüe that the commercial examined constitutes a "multimodal metaphor" (Forceville, 2004 and2005) through which the audience is essentially invited to interpret the Athens 2004 Olympic Games in terms of a festival; secondly, to explain how a commercial which foregrounds mainly festivities and celebrations is almost unanimously received by a multicultural audience as successfiílly advertising the Athens 2004 Olympic Games.Additionally, by performing an analysis of the mechanisms of the multimodal metaphor at hand, a general aim of our paper is to contribute to the study of non-verbal metaphor, an área much neglected by previous research on metaphor, as Forceville (2004Forceville ( , 2005) ) rightly points out.
The paper consists of two parts: the firstpart (sections 2-3) presents the commercial in question and discusses issues of multimodality and metaphoricity related to it, while the second part (sections 4-7) presents our hypotheses and questionnaire-based research, and discusses findings related to the interpretation and evaluation of the commercial by a sample of Greek and non-Greek viewers.

Advertising the Olympic Games 2004: A multimodal communicative act
Televised advertisements are instances of multimodal discourse, as, in Kress and van Leeuwen's (2001) terms, several semiotic modes are involved in their design, production and interpretation.The commercial examined here involves core visual modes 1 such as visual image, written language (subtitles and logo of the Games), gaze, movement (including running and dancing) and camera movement which, as Rheindorf (2004: 151) argües, has already been established as a semiotic mode in film.Core auditory modes involve spoken language (voice-over), music (instrumental), and sound.As we will argüe below (3.2), camera work and music play a significant role in the construction of the multimodal metaphor we examine.
In terms of content, the commercial at hand consists of nine thematic units.The first eight units show action-in-process which involves people celebrating some sports victory or simply having fun and enjoying themselves.Most of these units represent action that takes place outdoors.Some of the units are accompanied by (an unseen male' s) voice-over, while the words uttered are also shown at the bottom of the screen as subtitles.The commercial ends with a still frame in which the noun phrase' Olympic Games Athens 2004' and the symbol of these Games, a wreath made of olive-tree leaves, appear in white font against a sky-blue background.While this frame is shown, the noun phrase 'Olympic Games Athens 2004' is also heard in voice-over.
In order of appearance, the nine units of the commercial are: 1. Young people dancing in a club 2. A large group of men celebrating a Greek victory at a Greek coffee-house 3. A group of Brazilians dancing and celebrating in the streets of a Greek town 4. People in cars celebrating a Greek victory in the streets 5.A group of young people partying on a Mediterraneanbeach 6.A young man celebrating his medal victory at a Greek family meal 7. A fencing team celebrating victory in a locker room 8.A large crowd doing the Mexican wave in a stadium 9. Olympic Games Athens 2004 Table 1.Order of thematic units as shown in the commercial In order to analyse the commercial in the framework of tele-film studies, we used the following levéis of analysis:/rame, which is "a salient or representative still of a shot", shot, in which "the camera movement is unedited (uncut)" and scene, where "the camera remains in one time-space, but is at the same time madeup of more thanone shot" (Iedema, 2001: 141).Analysing the commercial in shots and scenes enriched our understanding of the thematic units presented in it and of the way in which visual images interact with sound and camera work in the metaphorical mapping of 'festival' features onto the 'Athens 2004 Olympic Games'.
Studies on multimodal interaction have suggested various elabórate frameworks for the analysis and transcription of multimodal acts that attempt to depict the many modes that interact with, and reinforce one another, for the production of meaning (see, for example, Norris, 2004, or Thibault, 2001).Below is a schematic presentation of the Athens 2004 commercial, which is our own adaptation from the extremely detailed and elabórate transcription of Thibault (2001).In the transcription of the commercial we have included a column listing scene numbers (which correspond to the nine thematic units outlined above), salient visual images/shots, movement (kinesic action), sound (voice-over and sound from the surrounding context), and music, in this case an instrumental tune thatplays throughout the commercial and dynamically shifts from being the most important sound in the scene (Figure) to being a contextual, taken-for-granted sound (Ground).The semiotic modes we have included in our transcription were the ones that we considered to be directly related to the construal of the multimodal metaphor involved in the commercial.In the above transcription, the column entitled 'kinesic action' presents a summary of the action depicted in the particular scene, whereas 'salient shots' are cióse shots from this scene that cover the whole screen and relate to viewers' positioning vis-á-vis the message.Sound involves voice-over as well as other surrounding sounds such as hooting or cries of joy.We should also note that music and voice-over altérnate in becoming the focal sound of most scenes.The transcribed scenes and shots presented here are the ones that are closely related to the construction of a multimodal metaphor through which the viewer of the commercial is invited to see the Athens 2004 Olympic Games as a festival of celebration and enjoyment.

Constructing and interpreting metaphors in multimodal communication
Our initial interpretation of the commercial at hand was that it involves a creative multimodal metaphor, one that invites its recipients to find similarities between two entities that have never been related to each other before.In other words, this is a metaphor for the identification of which it is not always possible to draw on shared ready-made categorisations and stereotypes.It is the kind of metaphor that Lakoff andJohnson (1989, in Forceville, 2005: 268) have termed 'one-shot metaphor' in order to distinguish it from conventionalized metaphors.In this section, we will attempt to substantiate our interpretation of the commercial by discussing it in terms of Black's (1979) 'interaction theory' of metaphor and Forceville's (1996Forceville's ( , 2004Forceville's ( and 2005) ) work on pictorial and multimodal metaphors.
In his 'interaction theory' of metaphor, Black (1979) focuses on novel and creative metaphors, such as the one discussed in this paper, and argües that innovative metaphors are intriguing in that they may point to a whole new way of thinking.Metaphors, in his terms, involve a target (primary subject) and a source domain (secondary subject) and any metaphorical text works by projecting upon the target domain selected features of the source domain.For example, in the "man is a wolf" metaphor, selected features of the source domain (wolf), such as' cruelty', are projected upon the target domain (man).In the section that follows we proceed to examine whether the multimodal phenomenon under examination constitutes a metaphor.
3.1.Are the Athens 2004 Olympic Games advertised through a metaphor?
Drawing analogies from Black's theory, Forceville (1996: 108) poses three questions as criteria for the identification of metaphors.In order for a phenomenon to qualify as a metaphor, there have to be clear and specific answers to all three questions.We will now present Forceville's questions and answer them as a way of providing evidence for our claim that what we are watching in this commercial is a metaphor.
(1) What are its two domains?
(2) What is its target domain, and what its source domain?
(3) Which feature or (structured) cluster of features can/must be mapped from source to target?
The first question is clearly answerable: the two domains are ATHENS 2004 OLYMPIC GAMES and FESTIVAL.Verbalising the metaphor is in this case relatively straightforward, as the two domains are explicitly referred to in scenes 1,7,8 and 9 of the commercial (see Table 2) 2 .The answer to the second question is equally clear, as ATHENS 2004 OLYMPIC GAMES is the target domain (this advertisement is part of the campaign promoting the event) and FESTIVAL the source domain, whose features are to be mapped onto the target domain.Some of the features that can be mapped onto the target domain are entertainment I merry-making, togetherness, victory 3 (which presupposes competitiori), celebration (that comes as a result of winning), and, overall,_/'oy.Through emphasising desirable and positive aspects of the source domain, the metaphor is possibly attempting to persuade the audience that watching the Athens 2004 Olympic Games will be a joyful experience, a chance to enjoy themselves and celébrate.
3.2.How is the particular metaphor construed?
According to Forceville (2004: 66), metaphors "whose target and source domains are predominantly or entirely presented in two or more different modes are 'multimodal metaphors'".As he (2005) rightly observes, in multimodal metaphors, target and source domain may be cued and identified in a non-linear, complex manner.In particular, he (2005: 274-5) argües that in metaphors in moving images "(1) target and source need not be presented at the same moment; (2) a domain (target or source) can be cued sonically as well as pictorially and/or verbally; (3) moving images allow for the establishment of similarity via camera angles and frames as well as via camera movements: two phenomena can be saliently emphasised using any of these stylistic options to créate, or help créate, metaphoric similarity".In this section, we will address the process through which the metaphor ATHENS 2004 OLYMPIC GAMES ARE A FESTIVAL can be identified by discussing the visual and aural means of its construal.
At first glance, it seems that the target and source domains of the metaphor examined in the televised advertisement are cued through interplay of the two core modes exploited by the particular médium, the visual and the auditory mode.In more specific terms, the metaphor is construed both visually (frames/pictures and written language) and aurally (voice-over, music), which confirms the multimodal nature of the metaphor.
The source domain of the metaphor is cued from the very beginning of the commercial both visually (people enjoying themselves in a dance club) and aurally (dance club music can be heard): entertainment and joy are two features of the first scene (dancing in a club), which are to be mapped onto the announced (by voice-over) 'festival' ("there is a festival...), which, however, is not further specified at this point.More features of the target domain are to be found in the scenes that follow, all cued visually and reinforced by music: scenes 2, 3, 4, 6, 7 and 8 foreground togetherness, celebration and joy, while, similarly to scene 1, scene 5 foregrounds partying and entertainment.Additionally, scenes 2, 6 and 7 foreground competition, victory and celebration.
Music interacting with quickly alternating shots is a crucial element of this multimodal metaphor: as an instrumental vivaciously beating tune that plays throughout the commercial, it provides the unifying link among seemingly unrelated scenes and henee, facilitates interpretation.In the case of the commercial examined here, same music in scenes 1-8 implies that the announced 'festival' (the Athens 2004 Olympic Games) will be joyful and fun like a party on the beach, or like dancing in a club; also, it probably implies that the Games will be a huge celebration, like the one we are watching in these scenes.In most scenes, music and sound (voice-over, hooting of cars, people's cries of joy) altérnate in the position of 'Figure' (van Leeuwen, 1999).More specifically, while the beating tune begins as clear Figure, as the most important sound of the first scene, it quickly loses this position to voice-over and is reduced to Ground, which is "still part of the listener's social world, but only in a minor and less involved way" (van Leeuwen, 1999: 23).The same pattern is repeated in the scenes that follow, with music returning as a foregrounded element in the next scene, and then becoming part of the familiar, taken-for-granted context.New sounds, such as hooting from cars or cries of people who celébrate become Figure, while the already familiar, instrumental tune, becomes Ground.Despite these changes in position, we would like to argüe that music remains a crucial mode to the construal of the metaphor, in that it helps viewers maintain the metaphorical interpretation of joy and celebration, necessary for understanding the commercial.
That the source domain is cued primarily by music confirms Forceville's (2004: 75) hypothesis that sound in a metaphor tends to cue its source rather than its target domain.Indeed, the beating tune that plays throughout the commercial cues the domain of festival and the ensuing connotations of enjoying one's self, celebrating and partying.Nowhere in the commercial does music cue the Athens 2004 Olympic Games in any way.Furthermore, our identification of the source domain through music and visual image confirms another of Forceville's (2004: 75) hypothesis, namely that sound alone is seldom responsible for cuing the source domain.We believe that cuing the source before the target domain is done in order to arouse and maintain the audience's interest, but more importantly, in order to trigger the audience's construction of the intended metaphor.
Interestingly, the target domain of the metaphor is not cued explicitly until the very last frame of the commercial, when the music stops and we hear (and also see written) the phrase ATHENS 2004 OLYMPIC GAMES.The target domain is thus construed verbally and pictorially, but not sonically, which is reserved for cuing the source.In our view, leaving the presentation of the target domain last is a successful means of maintaining the audience's interest throughout the commercial, as the target is only hinted atbut not fully revealed until the very last scene.Indications to the target domain are provided in the form of some salient shots which present clues concerning the nature of the festival (a Greek athlete winning a race in scene 2, a young medal winner showing his medal to the camera in scene 6) and the location of the event (the 'Brazilian' team dancing in front of a taverna called Zorbas in scene 3, or a Greek flag covering the sun in scene 4).Preceding the target domain with 8 scenes that cue the source, manages, in our view, to imprint in the minds of the audience the features of the source (see 3.1) that are to be mapped onto the target domain.
Our initial interpretation of the Athens 2004 commercial is fhus substantiated and confirmed: the commercial examined constitutes a multimodal metaphor, a novel and creative metaphor, which invites people to view the Athens 2004 Olympic Games as a festival, an occasion for competition, celebration, and entertainment.

Forming the hypotheses
As already argued, the Athens 2004 Olympic Games commercial constitutes a multimodal metaphor which invites viewers to relate a series of scenes to one another in order to identify the advertised product and arrive at the intended interpretation of the commercial.After having watched the commercial (scenes 1 to 9), one can affirm that some of the scenes presented in it are clearly more relevant to the advertised product than others.Table 3 presents the nine scenes in a hierarchy of growing relevance to the advertised product: top-down the numerical order represents a presumed least-to-most-relevant order in relation to the Athens 2004 Olympic Games.The actual but quite haphazard ordering of the scenes/thematic units (see Table 1) engages viewers in extensive inferencing in order to identify the implicit assumptions that relate all scenes to one another and to the advertised product.In this respect, Table 3 can also be seen as a guide to the inferential steps one presumably needs to follow in order to arrive at the intended interpretation of the multimodal metaphor in question and identify the advertised commodity (Athens 2004 Olympic Games).

Least relevant
Most relevant l.Clubbing 2. Partying on a Mediterranean beach 3. Doing the Mexican wave at the stadium 4. Celebrating victory in the locker room 5.A group of Brazilians celebrating in a Greek town 6. Celebrating medal victory at a Greek family meal 7. Celebrating Greek victory at a Greek coffee-house 8.People in cars celebrating a Greek victory in the streets 9. Olympic Games Athens 2004 Table 3. Order of relevant scenes to the Athens 2004 Olympic Games commercial Of the above scenes, the first two bear no relation to the Olympic Games in Greece, or to sports events in general.From 3 onwards, scenes gradually become more sport-related, but never clearly Olympic Games-related, or Greek Olympic Games-related.The relevance of scenes culminates with the explicit presentation of the Athens 2004 Games logo in the still frame at the end of the commercial.Inferential steps 1-8 show the extensive inferential process (from less relevant to more relevant stimuli) viewers have to engage in to arrive at the intended interpretation which is clearly retrieved (because explicitly prompted by the written logo) only upon viewing the final frame of the commercial.
Given the multimodal metaphor analysed above and the varying degrees of relevance of the scenes to the advertised product, the following hypotheses concerning the inferential process through which viewers attempt an interpretation of the commercial have been tested: 1. Regardless of their nationality, viewers will identify the relation of the commercial to sports and to the Athens 2004 Olympic Games in particular.2. Regardless of their nationality, viewers may describe parts of the commercial as not relevant to the advertised commodity.3. Due to 'irrelevant' scenes, viewers may admit to a certain degree of difficulty in retrieving the intended message.4. Regardless of their nationality, viewers will identify various shared implications of joy and entertainment (while attending sports events)

Method and design
In order to test the above hypotheses, a questionnaire was designed and distributed to two samplegroups, one consisting of Greek nationals (31 subjects) and the other of non-Greek nationals (28 subjects) from a variety of countries such as Albania, Bulgaria, Congo, Denmark, Germany, Hungary, Israel, Italy, Romania and Ukraine.Subjects were firstyear students of a language course in Modern Greek offered by the University of Athens and were recruited on a voluntary basis.
The questionnaire consisted of two parts: Part A involved identifying the advertised product and the general message of the commercial as well as estimating the overall difficulty encountered in doing so.Part B assessed the inferencing process in terms of specific aspects of the commercial (e.g. its culture-specific character) and the degree of difficulty viewers had when interpreting it.The two groups were given the questionnaire on sepárate days, but the process of filling it in was identical in both cases.In particular, each group first watched the commercial but its very last frame (unit 9), where the advertised product was explicitly shown on the screen, and answered questions in Part A. After having completed Part A, the group watched the commercial again, this time from beginning to end (last frame included), and answered the questions in Part B. The whole procedure (watching the commercial twice and answering the questions) lasted about 30 minutes.

Description of findings
As can be seen in Tables 4 and 5 (Appendix), native and non-native subjects' reactions to the TV commercial varied with respect to one issue out of eleven raised in the questionnaire, namely the cultural-specificity of the commercial (Tables 4 and 5, PartB4).Half of non-native viewers considered the TV commercial as multicultural and the rest of them were torn between a Greek-specific and a blend of Greek+international feature commercial.Native speakers reckoned the TV commercial was Greek-specific, with the minority torn between a multicultural and a blend of Greek+international feature commercial.Note however that, as will be shown next, the discrepancy between native and non-native audience's metacultural perspectives did not interfere with their successful interpretation of the televised promo.
Native and non-native subjects' reactions to the TV commercial converged with respect to the following issues raised in the questionnaire: (a) identifying the topic of the commercial as "publicizing a sports event" (Tables 4 and  5, PartAl) (b) describing the message of the commercial as "join in the joy of an International competition" (Tables 4 and 5, PartA2) (c) identifying an obvious relation to the Greek Olympic games (Tables 4 and 5, PartBó) (d) considering part of the material included in the TV commercial to be irrelevant (Tables 4and5,PartB7a) (e) retrieving a wide range of implied meanings (Tables 4 and 5, PartB8) (f) evaluating it as a successful TV commercial (Tables 4 and 5, PartB9) (g) easy understanding in terms of time, effort, background knowledge required (Tables 4and5,PartB10) The success of this commercial was confirmed by our subjects both explicitly, by answering 'yes, it was a successful ad' (78,57% non-native -93,54% native subjects, Tables 4 and 5, PartB9) and implicitly, by identifying a wide range of implicated assumptions (50% non-native -41,9% native subjects), which, irrespective of nationality, subjects seemed to largely share (Tables 4 and 5, PartB8 and Table 6, C).For the purposes of this analysis, the commercial was considered successful ontwo grounds: (a) if subjects identified the advertised product (Tables 4 and 5, PartAl) and (b) if subjects identified a range of relevant shared contextual implications (Table 6, C).As for uptake effects, both groups acknowledged little difficulty in retrieving the intended interpretation, which disconfirmed our initial hypothesis 3 (non-native 28,57%-native subj ects 29,03 %) (Tables 4 and 5, PartA3 and PartBIO).Note that the central, for the purposes of this study, issue of difficulty in identifying the message communicated was raised twice: firstly, inPartA3, aiming at testing viewers' difficulty in understanding the core meaning, which we considered to be automatically retrievable and henee, explicit; secondly, in PartBIO, aiming at testing viewers' awareness of specific 'difficulty' factors (time, mental effort, background knowledge required) upon reflection.In the latter case, 'difficulty' referred to identifying not only the explicit but any implicit assumptions delivered too (PartB8).Viewers' unanimous success in interpreting the televised commercial (as intended by the advertiser) was largely expected (see Hypotheses).What has hardly been anticipated though was the viewers' reaction to issues (e) and (g) above, i.e. the type of implicatures they retrieved (varying/shared, weak/strong) and the correlation they established between 'difficulty-to-understand' with 'time-effort-background knowledge required to do so'.In fact, it is the viewers' ability to derive a wide range of shared implicatures and to intuitively counterbalance 'easy understanding' with 'extra time, effort, background knowledge required' that seem to contradict standard assumptions in the analysis of metaphorical advertisements (see Forceville, 1996).
With reference to issue (e), native and non-native viewers spelled out a considerable number of implicit assumptions the commercial conveyed (see Table 6, C), which quite interestingly, roughly coincided with the core message viewers retrieved (Table 6, A) before exposed to the explicit logo (final frame, scene 9).If A (Table 6) is taken to be the explicit meaning communicated and C (Table 6) the implicit meaning communicated, and if the two overlap, as viewers' responses indicated, then our hypothesis that viewers retrieved a range of shared and henee, strong implicatures, instead of an explicit meaning distinct from many weakly implicated ones, seems to be strengthened.Contrary to standard cases of one-to-one verbal communication, where the explicit meaning is clearly distinguished from a range of implicatures that may or may not be retrieved, addressing a large audience for promoting an international event while aiming at the highest possible publicity seems to be exclusively relying on strongly communicated implicatures.In this respect, the results of testing the Athens 2004 OG multimodal televised advertisement contradict the analysis of printed advertisements as discussed by Forceville (1996) or Tanaka (1994) in a number of respects.Firstly, Forceville's assumption that "communicator and addressee share a linguistic and cultural background" (ibid.: 96) does not hold for our multi-ethnic sample: the linguistic and cultural mismatch between Greek advertiser and non-Greek target group did not seem to interfere with a fruitful interpretation process despite the cultural variability brought out in viewers' answers (Tables 4 and 5,  PartB4).Secondly, Forceville's claim that metaphors only achieve relevance through a wide array of idiosyncratic weak implicatures (1996: 97), which are weak precisely because they vary from individual to individual (Forceville, 1996: 102), does not seem to be borne out by the evidence.The advertiser attempts to actívate, in practice, some kind of common denominator in her target audience's cognitive environments.If anything is communicated beyond that, it will vary from individual to individual and is therefore, I propose, weakly communicated (ibid.).
It seems to follow that Forceville rejects the possibility of a "common denominator" of assumptions that is shared by and is strongly communicated to a large audience.Yet, our Greek and non-Greek audiences recovered strikingly homogeneous assumptions as expressed under PartAl and PartB8 (see Appendix, Table 6 A, C).Thirdly, Forceville claims that "text is a vital component in making sense of an ad" so that "the less verbal anchoring is present, ... the greater the range of interpretations that is possible" (1996: 102).But the limited Greek linguistic input (subtitles/voice-over) in the present multimodal TV commercial was beyond non-Greek viewers linguistic ability 4 , still they interpreted the commercial along the Unes Greek viewers did, retrieved the same or similar assumptions, rated it as a successful advertisement and more importantly, considered it to be easy to understand.In fact, we believe that in the absence of any constraining linguistic input, the same multimodal advertisement drawing exclusively on its visual and acoustic input (music, colour, movement, gesture) would trigger the same or possibly more implicatures on viewers' part regardless of nationality.In the absence of any text, interpretation depends, of course, heavily on inference processes; an explanation of how extensive inferencing (in terms of time, effort, background knowledge) is offset by a wide range of implicatures is offered while discussing issue (g).Finally, contrary to standard commercial advertisements which are, according to Tanaka (1994), covertly communicated (because the advertiser wants to keep his intention to influence the viewers covert), the televised promo under discussion deviates from the norm by being a case of overt communication.This is so because the aim in publicizing the Olympic Games was other than directly promoting a product to sell; the aim was to invite Greek people (possibly non-Greek residents too) to contribute to the hospitable, joyful, cooperative, amicable and volunteering spirit of the Games so that the 2004 OG in Athens runs smoofhly and sells well.There seems to be no obvious motive for communicating such a message covertly; on the contrary, there is a strong motive for publicizing the advertiser's informative intention: to influence the audience towards a good cause 5 .
With reference to issue (g), viewers' extensive comments on the last open-ended question wereparticularly intriguing, since the majority (71,42 %) answered that it was an msy-to-understand commercial but pointed out that it required some time, some effort and some background knowledge to access the intended interpretation.This prima facie puzzling observation by both native and non-native viewers deserves an explanation, if only for the oxymoron themselves acknowledged in an overall effective interpretation process.
The main question addressed in the next section refers to how we can explain the fact that native and non-native speakers easily inferred the target domain of the metaphor and unanimously considered the commercial to be a successftil TV ad despite the unanimously acknowledged 'irrelevant' scenes and despite the different cultural group-perspectives: non-native viewers considered the TV commercial as multicultural whereas native viewers considered it as Greek-specific.Using the framework of Relevance theory (Sperber and Wilson, 1986/95;Wilson and Sperber, 2004), an explanation is sketched of how it is that a culturally-specific TV ad drawing on two seemingly disparate themes -sports and joyful life -is a successful means of multimodal communication, as shown in the findings of the research project described in this paper (for an overview of the findings, see Appendix, Chart 1).

Explanation of findings
Relevance theory originated as a post-Gricean pragmatic account of utterance interpretation.How hearers bridge the gap between sentence meaning and speaker meaning 6 rests on a cognitively-driven account and the claim that human cognition is geared towards the maximization of relevance.For Sperber and Wilson (1986/95), 'relevance' is generally characterized as a property of inputs -external stimuli or internal representations -to cognitive processes.An input may bear different types of cognitive effect -it may combine inferentially with existing assumptions to yield new conclusions (contextual implications), it may contribute new evidence that confirms existing assumptions, or it may contradict and eliminate already known information.For Sperber and Wilson, relevance is not exhausted by the effects an input might yield; relevance is a delicate balancing act between cognitive effects and processing effort: the greater the ratio of effects to effort the greater the relevance of an input.Under this definition of relevance, then, for humans to aim at maximizing relevance is to aim at retrieving as many positive cognitive effects as possible for as little processing effort as possible.The basic tenets of relevance theory appear below (see Wilson and Sperber, 2004).
Relevance of an input to an individual a.Other things being equal, the greater the positive cognitive effects achieved by processing an input, the greater the relevance of the input to the individual at that time.
b.Other things being equal, the greater the processing effort expended, the lower the relevance of the input to the individual at that time.

Cognitive Principie of Relevance
Human cognition tends to be geared to the maximisation of relevance.

Communicative Principie of Relevance
Every ostensive stimulus conveys a presumption of its own optimal relevance.

Optimal relevance
An ostensive stimulus is optimally relevant to an audience iff: (a) It is relevant enough to be worth the audience's processing effort; (b) It is the most relevant one compatible wifh communicator's abilities and preferences (or goals).
Relevance-theoretic comprehension procedure a. Follow a path of least effort in computing cognitive effects: Test interpretive hypofheses (disambiguations, reference resolutions, implicatures, etc.) in order of accessibility.b.Stop when your expectations of relevance are satisfied.
These tenets are couched upon the assumption that, not only linguistic, but all communication is relevance-driven.However, human verbal communication has been so far the main test-bed for relevance-theoretic ideas (Sperber and Wilson, 1986/95;Wilson and Sperber, 2004).With few exceptions (see Forceville 2004;Yus 1998), existing relevance-theoretic work on multimodal communication involves primarily static verbo-jjictoriaZ representations (see Forceville, 1996 andTanaka, 1994onadvertisements).This work shifts the emphasis towards the audience's reaction to an animated multimodal communicative stimulus -the 'Athens 2004 Olympic Games' TV commercial -featuring moving pictures, music, sound/visual effects and spoken/written language in certain scenes only (see Table 2) -and to whether the predictions relevance theory makes for the audience's reaction are confirmed or disconfirmed through empirical evidence.More importantly, the verbal element is suspended in the case of our non-native viewers, which turns this into a genuine non-linguistic experiment assessing viewers' reaction to an exclusively non-verbal multimodal stimulus.
Turning to the results of this work, the fact that non-native viewers retrieved an intemational reading of the ad follows directly from the cognitive principie of relevance and the relevance-theoretic comprehension procedure.In their effort to maximize the relevance of the promo, non-native viewers followed the path of least effort by drawing primarily on the intemational scenes (Brazilian football funs, club dancing, fencers and champagne uncorking in the locker-rooms, Mexican wave) and combined them with the culturallyspecific Greek scenes (family celebration for grandson's medal win, traditional coffee house celebration for Greek athlete's victory, Euro football 2004 celebration for Greek championship) to stop as soon as they retrieved the 'enjoy the Olympic Games in Greece' interpretation.Similarly, the fact that native speakers retrieved a Greek reading of the ad follows directly from the cognitive principie of relevance and the relevance-theoretic comprehension procedure.Under a default tendency to process maximally relevant information, native speakers followed the path of least effort by easily tuning in to the familiar Greek scenes (traditional coffee-house, 'Zorbas' taverna, Greek flag, large family meal) and combined them with the internationally-oriented scenes to stop as soon as they retrieved the 'enjoy the Olympic Games in Greece' interpretation.
Interestingly, however, both native and non-native speakers considered many of the scenes to be irrelevant to the core message and more effortful to process in terms of time, mental effort and background knowledge required.This is predicted by the characterization of relevance in terms of effect and effort (relevance ofan input to an individual): according to clause (b), the greater the processing effort (scenes 1-8, Table 3), the lower the relevance of the commercial -henee viewers' reckoning of irrelevant scenes and some (more than expected or desired) effort or time or background knowledge to get at the intended interpretation; according to clause (a), the greater the positive cognitive effeets achieved by processing the televised promo (peaking at scene 9, Table 3), the greater its relevance -henee, viewers' reckoning of 'irrelevant' scenes as an essential guide to retrieving the intended meaning.
The fact that the promo designer did not provide the most economical and least effortful Olympic Games logo with the explicit script 'Join in and enjoy the Olympic Games 2004 in Athens' (scene 9) in the first place, but engaged viewers in additional processing effort (scenes 1-8) instead, which demands extra inferencing in terms of accessing background sociocultural practices and conventions, is bound to be offset by a rich enough set of positive cognitive effeets.A good communicator's choice of the message-format to use is determined not necessarily by what is the least effortful or most economical but by her "judging more or less correctly what a hearer can easily infer in a particular context" (Carston, 2005: 276).As Carston shrewdly observes (ibid.), the role of the speaker (or communicator) is not to over-encode her meaning(s) by being tediously, pedantically or misleadingly explicit.According to the key theoretical concept of 'optimal relevance', the communicator aims at producing a message that is optimally relevant to the addressee but may fail to produce the most relevant one due to limitations in his abilities, preferences or goals.A communicator often prefers to risk conveying additional, insinuating information (scenes 1-8) rather than employing the most economical format (scene 9) to communicate her message in order to créate a more impressing, more memorable, more forceful effect instead of a less roundabout but dull one.
Surprisingly, despite the seemingly irrelevant scenes and the worth-mentioning effort, viewers derived a wide range of specific implicit assumptions these scenes aimed at conveying.This is predicted by the Communicative Principie of Relevance according to which every communicative act conveys a precise presumption of its own optimal relevance.According to clause (a) of this definition, the addressee is entitled to expect her communicative input to be at least relevant enough to be worth processing and (given the Cognitive Principie of Relévame) more relevant than any altemative input available at the time.This minimum expectation of relevance follows from clause (b), which spells out the communicator's (here, the ad creator's) concern to make her product as easy as possible to understand with the greatest cognitive effects.But communicators are, as already suggested, restricted by their own abilities, preferences or goals.Henee, some aspect(s) of the information conveyed may strike the audience as less appealing, less relevant or less easily understood.Still, because the very act of communication raises an expectation of its own optimal relevance (Communicative Principie of Relevance), hearers or audiences follow a specific inferential procedure, the relevance-theoretic comprehensionprocedure, to get at what the communicator intended to convey, or else, to achieve greatest positive cognitive effects in a cost-effective way.

Conclusions
The findings of the experimental work reported in this paper on how a multimodal televised promo is interpreted by a multicultural audience vs. a native audience can be summarized in the following observations.Multimodal metaphors may convey specific strong implicatures that audiences unanimously reckon, not necessarily a range of weak implicatures (Forceville, 1996).In the framework of Relevance theory, the multimodal metaphor discussed in this paper is shown to be cross-culturally effective because it draws on a trade-off between 'least effort' and 'satisfying one's expectations of relevance' (as determined by the relevance-theoretic comprehension procedure).This multimodal metaphor involves prima facie non-relevant information which, nonetheless, subjeets evaluated as critically contributing to deriving the intended interpretation.We have suggested that the apparent conflict between' irrelevant scenes' and' correct identification of intended meaning due to 'irrelevant' scenes' is offset by a trade-off between positive cognitive effects and processing effort triggered by a cost-effective inferential procedure (as determined by the definition of relevance ofan input to an individual).
Notes * The authors wish to thank Sophia Marmaridou, Mary Sifianou and the audiences of the 2005 IPrA conference and the 2005 HASE conference for interesting comments on earlier versions of this paper as well as the editors of RAEI and the anonymous reviewers for kindly offering their improving suggestions.
1.The categories of semiotic modes presented here are based on Rheindorf' s (2004:146) map of semiotic modes in film.Although his map is not as detailed as that of others (see, for example, Stockl's (2004) network of modes, sub-modes and features in TV-and film-media), it was selected as the most appropriate for the purposes of our analysis.
2. The difficulty we encountered with verbalising the two domains of the metaphor related to the translation of the Greek word yiopTfj (celebration, holiday, feast, féte, festival, festivity), in that we needed a term whose meaning would encompass the notions of joyful gathering, competition and entertainment.In this respect, the word 'festival' was deemed more appropriate.
3. The presentadon of winning athletes and teams celebrating their victory is likely to convey implicit meanings (atleast, in the mindsof the Greek viewers), namely that the organisation of the Olympic Games will be successful, and/or that Greek athletes will be victorious.
4. Non-native viewers were attending the University of Athens beginners' course in Modern Greek language.Their minimal linguistic competence in Modern Greek became manifest in their failure to follow our explanatory remarks in Greek (aimed to familiarize them with the experimental procedure), in which case we had to shift to English.
5. The Athens 2004 Olympic Games organizers' had openly declared their intention to appeal to people's collaborative spirit and polite manners through a number of other, more explicit televised messages, e.g. by suggesting how stadiums should be kept clean, or how visitors should be treated in restaurants, taxis, sight-seeing tours.As a consequence, the advertiser's overt intention to communicate an invitation along the lines of 'join in the joy and success of the Athens 2004 Olympic Games' is, for the purposes of this analysis, given.
6.In the contextualist approach to semantics (see Sperber and Wilson 2005), sentence meaning is semantically incomplete and fragmentary meaning that the hearer must inferentially enrich at the explicit level (well beyond disambiguation and reference assignment).Speaker's meaning is the enriched sentence meaning that the hearer will further pragmatically elabórate on at the implicit level "by supplying contextual assumptions which will combine with it (the enriched sentence meaning) to yield enough conclusions (or other cognitive effects) to make the utterance relevant in the expected way" (ibid.: 361). Figure Humanity united in friendly, joyful competition To attract holidaying visitors and celébrate sports united in a partying atmosphere To motivate partícipation, promote Greek holidaying, lifestyle and way of celebrating sports events in a partying atmosphere Enjoy the spirit of the games Be part of the games Have great holidays Greeks are multicultural and hospitable Celébrate, have fun united Attend, support the Olympic games Greece of joy and entertainmentTable 6. Unanimous native and non-native subjeets' reaction to TV commercial with respect to open-ended questions.