A Multimodal Discourse Analysis of the Interactive Meaning in Public Service Advertisement

On the basis of Kress and van Leeuwen’s Visual Grammar based on Halliday’s Systemic Functional Linguistics, this study explores the interactive meaning in three public service advertisement multimodal discourses, adding evidence to the assumption that Systemic Functional Linguistics can be applied to the multimodal discourse analysis of public service advertisement in a feasible and operational manner.


Introduction
Nowadays, advertisements are increasingly penetrating into people's daily life, playing an increasingly important role. Whether through traditional media tools, such as newspapers, books, letters, or modern media tools, such as television, computers, mobile phones or other network terminals, advertisements can be found anywhere at any time, as if advertising has become one of the indispensable elements of human society. Generally speaking, advertisement can be divided into two categories, namely, commercial advertisement and public service advertisement (PSA). The criteria for classifying the two types of advertisements are also quite simple, that is, whether they are profit-oriented or not. From research in recent years, it can be seen that experts and scholars on advertising intend to set their research focus on commercial advertising, while research on PSA is comparatively less. In addition, research perspectives need to be expanded in an all-round way, and research connotations need to be further explored.
With information technology rapidly developing, visual language becomes more and more important in all aspects, especially in advertising. The multimodal feature of advertising is an evidence of its use of various semiotic resources, such as language, image, sound, and color, to better convey the connotation and obtain the best advertising effect. Multimodal discourses exist widely and experts have carried out a lot of research on the interpretation of multimodal discourses. Among these studies, multimodal discourse analysis (MDA) has taken a place, taking Systemic Functional Linguistics (SFL) as its theoretical basis and providing new research ideas and perspectives for discourse analysis. As a new research method of discourse analysis, MDA has become a research focus of western scholars in recent years, especially in the fields of linguistics and social semiotics. Kress and van Leeuwen [1,2], Lemke [3], Royce [4], O 'Halloran [5,6] and Baldry [7] all made positive and significant contributions to MDA. Chinese linguistics Li [8], Hu [9] and Zhu [10] introduced the theory of MDA. Nowadays, with the gradual development of MDA, more and more researchers are focusing on this field. However, there is a lack of MDA focusing on PSA, which leaves a lot of gaps to be filled.

SFG, VG, MDA and PSA
In 1990s, Halliday's SFL theory started to be applied in the study of MDA. Among the scholars who applied SFL to MDA, Kress and van Leeuwen [1] put forward a theory that becomes prevailing in the field -they proposed Visual Grammar (VG) to analyze visual images. On the other hand, studies on discourse analysis starts from 1950s attracting wide attention from scholars and researchers.

Systemic Functional Grammar (SFG)
As mentioned previously, much research conducted in the field of MDA are based on the theory of social semiotics, which in its narrow sense, refers to the Systemic Functional Grammar (SFG) of Halliday. The tenet of SFG is to regard language as a kind of social semiotic and study language from the perspective of its interaction with the social context. According to Eggins [11], SFG "has been applied to interpreting the grammar of other semiotic modes, such as visuals [1,2], sound [12] and action [13].
As Halliday [14] indicates, "In a functional grammar, a language is interpreted as a system of meanings, accompanied by forms through which the meanings can be realized". Therefore it can be said that SFG focuses on the means of realizing the functions of language. According to Halliday's framework, in language structure and human communication there are three meta-functions, namely, ideational function, interpersonal function and textual function. 1) Ideational Function: Language is used for describing, reflecting or interpreting the world around us, the individual experience of ourselves, consisting of "goings-on, happening, doing, sensing, meaning, and being and becoming" [14].
2) Interpersonal Function: Language is used for having communication or interaction with people, establishing and maintaining relations with them, influencing other people's behavior, expressing viewpoints of our own on the things in the world, and eliciting or changing other people's points of view. In this way, language is used to enable people to participate in communications, take own their roles, and express their own and understand others' feelings, attitude and judgements.
3) Textual Function: It refers to the fact that language itself has a mechanism to turn a random list of sentence into a living passage by organizing any stretch of written or spoken discourse into a coherent and unified text [14]. Therefore, in textual function, language is used to relate what is said or written to the real world or other linguistic events.

Visual Grammar (VG)
Inspired by Halliday's ideas of social semiotics, Kress and van Leeuwen proposed and developed a method of social semiotic analysis of visual communication and established a descriptive framework of multimodality. This theory of Kress and van Leeuwen carried out for reading images is called Visual Grammar (VG). In light of Halliday's three meta-functions of language -ideational, interpersonal and textual meanings, they put forward three meta-functions of images -representational, interactive and compositional meanings. The corresponding relation between the three meta-functions in SFL and VG is shown in Table 1 [15].

Ideational Representational
Interpersonal Interactive Textual Compositional Table 1 Three strands of meta-functions in SFL and VG 1) Representational Meaning: "Any semiotic system has to be able to represent, in a referential or pseudoreferential sense, aspects of the experiential world outside its particular system of signs" [1]. That is to say, the semiotic system has to be able to represent objects and their relations in the world outside the representational system. According to SFL, language "represents and constructs our perception of reality in the form of 'goings-on' or processes of various types, which incorporate different categories of participants (e.g. actors, goals, receivers, sensors, attributes)" [15].Adapting this model to the analysis of images, Kress and van Leeuwen [1] divide representational meaning into two main categories of processes: narrative and conceptual.
2) Compositional Meaning: "Any semiotic system has to have the capacity to form texts, complexes of signs which cohere both internally and with the context in and for which they were produced" [1]. The visual and verbal elements in multimodal discourses form a compositional whole. When expressing the textual metafunction, compositional resources would appear to be: (1) horizontal structure when presenting visual information as Given or New and (2) vertical structure when presenting visual information as Ideal and Real [1].
3) Interactive Meaning: "Any semiotic system has to be able to project the relations between the producer of a sign or complex sign, and the receiver/ reproducer of that sign" [1]. That is to say, any semiotic system has to be able to project a particular social relation between the producer, the viewer and the object represented.
According to Kress & van Leeuwen [1], there are two kinds of participants involved in an image, namely, represented participants (the people, the places and things depicted in images) and interactive participants (the people who communicate with each other through images, namely the producer and the viewers). Between these participants, there are three kinds of relations. First, relations between represented participants; Second, relations between interactive and represented participants, which refers to the interactive participants' attitudes towards the represented participants; And third, relations between interactive participants, which refers to the things interactive participants do to or for each other through images.
Interaction happens whenever the producers who produce the images and the viewers who appreciate the images expressed their own ideas on the ways of images interpretation. Some of this interaction is direct and immediate where the two parties have conversations face to face. While other interactions may be neither direct nor immediate due to the absence of both sides. When examining an interaction, four key factors shall be taken into careful consideration: visual contact, social distance between the represented participants and the viewers, perspective and modality accounting for the degree of factuality and truthfulness of visual images with regard to the real world [1].

Multimodal Discourse Analysis (MDA)
According to Xin [16], there are many well-known researchers studying MDA: O'Toole, Kress, van Leeuwen, Lemke, O'Halloran, Thibault and Ventola who studies MDA from the perspective of social semiotics; Scolion, Norris and Jewitt from the perspective of interactive sociology; Forceville and Holsanova from the perspective of cognitive. Kress and van Leeuwen's [1] framework documents the system of meaning for the ideational, interpersonal and textual which are respectively labeled as representational, interactive and compositional in visual communication. Reading Images explores how figures, places and things in the image compose the visual design to realize the compositional, interactive and representative meanings. As cited from Li (2003), many multimodal discourse analysts made great contributions to MDA. For example, O'Toole [17] analyzes images from the viewpoint of stratification; Lemke [3] explores the interrelations between the graphics and the verbiage in scientific articles. Royce [4] explores the image-verbiage relations in advertisements. In analyzing threedimensional discourses, a few analysts have done tentative studies. For example, Martinec [13] makes a multimodal study of body movements and the space design of the play.
In comparison with the prosperous studies on MDA overseas, China has just witnessed a budding stage in recent years. Several Chinese scholars have realized the importance of multimodality. A series of introductory papers to multimodality [8,9,10,16,18,19,20] and studies which apply the theories of MDA in specific types of discourses [21,22,23,24] have been done. Li [8] is the first to introduce the theory of MDA and reviews Kress and van Leeuwen's VG in her article. Afterwards Hu [9], Zhu [10], Xin [16] and Zhang [20] propose their opinions on MDA, systemically introducing the theories of MDA and outlining the research status, the limitations and the future of MDA. As for the application of the theory, plenty of research has been done. Hu and Dong [25] makes a multimodal analysis of PPT; Ding [26] probes the modality in the illustrations in fairy tales. Several researchers such as Li [27], Wang [23] and Hu [9] make a multimodal analysis of print advertisements and they mainly focus on the interaction between the image and the verbiage.

Public Service Advertisement (PSA)
Public Service Advertisement is to diffuse beneficial social values to the public by using diverse artistic demonstrational styles which can easily attract audience's attention, affecting their views and behavior towards the society and get a better solution to social problems. There are plentiful studies on advertising discourse in social context in the past few decades. Leech [28] first studied advertising discourses from the view of stylistics. Cook [29] put an emphasis of advertising discourse on the social implication of advertising language. One of the Chinese frontiers, Huang [30], applies Systemic-Functional Grammar in advertising analysis and gives a description of advertising from some aspects. However, they lack systemic analysis of inter-semiotic grammatical connections in PSA. Therefore, this thesis chooses PSA as data to explore the interactive meaning by using MDA.

Methodology
The purpose of this study is to analyze how the multimodal discourses of PSA interact and exert the effects. Thus, the present study attempts to provide an answer to the following questions: This study will analyze the visual and verbal messages in the PSAs respectively and then study the interaction relations and effects of the verbal and visual modes. Data analysis is composed of three sections. The first section is concerned with the interactive meaning in terms of the four elements of VG, namely, visual contact, social distance, perspective and modality; the second section analyzes how the interactive meaning is constructed in verbal language with respect to mood system and attitude system. The last part explores how the visual modes and the verbal modes cooperate to construct the interactive meaning in PSA multimodal discourse.

Discussions
In this section, three PSA examples will be analyzed respectively in detail in order to inquire how various modalities work together to realize the interactive meaning and to persuade the public in advertising discourses.

Analysis on the Interactive Meaning in the Visual Image
(1) Visual contact In terms of visual contact, this image can be counted as both a "demand image" and an "offer image" because that some of the represented, the two big polar bear, look directly at the viewers, while others, the two small polar bear and the penguin, do not look directly at the viewers. On one hand, "the participant's gaze demands something from the viewers, demands that the viewers enter into some kind of imaginary relation with him or her" [1]. That is to say, the represented participant wants the viewers to take some action or to establish some kind of social rapport with the viewers. In this case, the represented, polar bear and penguin standing on isolated ice, want the viewers, human, to connect the light turning on with the severe conditions of their habitats and then come up with solutions to ease the issue to protect these endangered animals. On the other hand, the small polar bear and penguin do not look directly at the viewers but gaze somewhere else. According to Kress and van Leeuwen [1], this kind of image "offers the represented participants to the viewers as items of information, objects of contemplation, impersonally, as though they specimens in a display case". In this case, the represented, as objects of contemplation, provides information to the viewers. Viewers should make use of this information, which is that the earth's climate is undergoing severe changes, and do some thinking on taking measures to save the animals.
(2) Social distance The represented in this image could be divided into two parts, namely, the light and the picture with animals on the wall. In terms of the light, a long shot is employed with the object shown out of the viewer's reach with much background information, which suggests a far social distance; while in terms of the animals, a very long shot is employed with a total of eight torsos shown to the viewers, which entails a public distance. Both the long shot and the very long shot indicate a rather far distance, which gives the viewers a feeling of objectivity. That is to say, the viewers are more easily convinced of the truthfulness of the advertisement and thus earnestly take actions to solve the objective problem. (

3) Perspective
For the horizontal dimension, a frontal angel is adopted to depict both the picture and the light. With a frontal angle, the image transmits involvement, indicating that climate issue and earth-protecting mission are worldwide and in need of every one's efforts and contributions. For the vertical angle, an eye-level angel is chosen to transmit equality between the represented and the viewers, showing that the viewers are all in the same boat when facing the issue shown in the image.
(3) Modality The degree of modality of an image is assessed from the following aspects: color, contextualization, representation, depth, illumination and brightness. This advertisement is truthful and reliable for the reason that it has high modality. For contextualization and representation, much details of the background as well as the represented participants are depicted, for instance, the wallpaper pattern on the wall, the small pieces of ice, the frame of the picture and the posture of the penguin. For color, the image has fully saturated colors, several colors and different shades of colors. In addition, there are also the play of light and shade as well as different degrees of brightness, especially areas around the lighting lamp.

Analysis on the Interactive Meaning in the Verbal Language
The verbal components identified in this PSA mainly include this one sentence on the upper side of the image: (1)"When you leave the light on, you are not the only one who pays" with the words "light on" in bold. The words (2)"panda.org/climate" on the right lower corner and the sign beside indicate the institution issuing this advertisement which is "WWF (World Wildlife Fund)" and its website. The verbal text (1) is a declarative sentence to give a moderate description on the consequences of not saving the energy, which is to make climate warming worse and leave some animals in an endangered situation. What is particular about this sentence is that it intends to convey the idea in an implicit but not explicit way, leaving spacious room for contemplation to the readers. The latter part of the sentence "you are not the only one who pays" intrigues the readers to reflect on the question "who else will pay" and may come up with answers themselves.

Analysis on the Visual-Verbal Inter-Semiotic Interaction
The visual and the verbal text of the advertisement cooperate to construct the interactive meaning of this image.
To begin with, in terms of visual contact, this image is a "demand image" and an "offer image", which demands people to take immediate actions and have deep contemplation by making direct visual contact and offering information of the worsening climate change; the far social distance in this image delivers objectivity; involvement and equality is conveyed through the perspective of the image; and this image is of high modality. As for the verbal text of the image, a declarative sentence is used to lead the viewers into thinking about the consequences of not being environmentally friendly. Therefore, the visual and the verbal text of the image integrate to address the topic issue -environmental protection and wildlife protection and to urge the human society to take actions to save the planet by doing everyone's own bit.

Analysis on the Interactive Meaning in the Visual Image
(1) Visual contact In terms of visual contact, this image can be counted as an "offer image" because the represented, the child crying heart-brokenly against the wall with his limbs of serious injury, do not make direct eye contact with the viewers. Therefore, the image offers the represented participants to the viewers as an item of information or an object of contemplation. The viewers shall consider the meaning this image intends to convey carefully to explore the messages contained.
(2) Social distance The represented in this image is depicted using a long shot for the reason that the viewers can see the whole human figure with space around it. By employing a long shot, a far social distance is suggested. At far social distance, the phenomenon in the image is presented in an objectively and impersonally manner. With the long shot, the image presents the objects for the viewer's contemplation, leading the viewers to think about the possible reasons behind such phenomenon as well as to find out effective and practical solutions to such issues.

(3) Perspective
For the horizontal dimension, an oblique angel is adopted to depict the child crying his heart out in the room. The oblique angle detaches the viewers from the represented, correspondent with the detachment conveyed by the far social distance between the represented participant and the viewers. For the vertical dimension, a low angle is adopted since the viewers are requested to look up to the represented participant. The low angle is suggestive of the importance of the portrayed participant and the significance of having the problem solved.

(4) Modality
This advertisement has high modality in terms of contextualization and representation, much details of the background as well as the represented participants are depicted, such as the door and the bolt and the lock on the door, the connector plugs on the wall, the doorsteps, the tiles on the floor, the broken pieces of the child, the facial expression of the child and so on. For color, the image has several colors and different shades of colors. In addition, there are also the play of light and shade such as the changing of light on the doorstep.

Analysis on the Interactive Meaning in the Verbal Language
The verbal components identified in this PSA mainly include this one sentence on the upper right side of the image: (1)"YOU CAN LOSE MORE THAN YOUR PATIENCE" with all the letters capitalized. On the lower right corner of the image is the sign of the institution who issued this advertisement and its website (2)"www.goodparent.pl". The letters in verbal text (1) are all in uppercase, which shows the producer's emphasis on the issue. This sentence brings the viewers to contemplate on the question "what else will I lose if I lose my patience" and they may come up with unexpected answers themselves. From the name of the website, readers may assume that this advertisement is to teach parents how to be good parents. And in this case, this advertisement is to teach parents to be patient to their children.

Analysis on the Visual-Verbal Inter-Semiotic Interaction
The visual and the verbal text of the advertisement cooperate to construct the interactive meaning of this image.
To begin with, in terms of visual contact, this image is an "offer image", which offers the viewers information for contemplation; the far social distance in this image delivers objectivity; an oblique and low angle is adopted in terms of perspective; and this image is of high modality. As for the verbal text of the image, a declarative sentence is used to lead the viewers into thinking about the consequences of not being patient to the children. Therefore, the visual and the verbal text of the image integrate to address the topic issue -parenting and child education and to urge parents to have contemplation on their loss-of-patience issue to their children. In this image, the represented object do not have their eyes depicted, not to mention whether they are looking directly at the viewers. In this case, the viewers may take it as both a "demand image" and an "offer image", or consider it as neither a "demand image" nor an "offer image", which depends on the viewers' imagination and association. As far as the author is concerned, the represented in the first image, the persona sitting on the bed, may make direct eye contact with the viewers; while the two persons sitting face to face at a dinner may not look directly at the viewers because it seems like they are having a conversation so they may just look at each other. In this way, both the "demand" and the "offer" images are shown to the viewers, leaving the viewers to intrigue their imagination and take relevant practices to the issue addressed.

Analysis on Example 3 "Cigarettes Smoke People"
(2) Social distance In this image, there are two parts when referring to the represented. The first part of the represented, the persons with cigarettes in their fingers, is depicted in the medium long shot since the full figure is shown, entailing a close social distance with the viewers. The second part of the represented, the cigarettes being hold in persons' hands, is depicted in the long shot since the object is shown out of the viewer's reach with much background information. The object is there for the viewer's contemplation. In this case, an invisible barrier exists between the viewer and the object possibly due to the extreme harm of the object -cigarettes can do harm to people's health as the image shows.
(3) Perspective For the horizontal dimension, an oblique angel is adopted to depict the represented, both the people holding cigarettes and the cigarettes. The oblique angel suggests that there is detachment rather than involvement between the represented and the viewers. For the vertical dimension, a high angle is chosen for the viewers may need to look down on the represented participants, which suggests the inferiority of the represented participants in terms of health conditions for the reason that cigarettes are doing harm to their physical quality.

(4) Modality
This image is highly truthful for it has moderately high modality. To begin with, this image has high modality in terms of contextualization and representation. A great deal of details of the background and the presented participants are depicted, for example, the jewelries on the women's hands, the smoke of the cigarettes, the lighting lamps on the both sides of the bed, the ashtray and phone on the night table, the coffee cups and sugar containers on the table, the lipstick on the brink of the cup, etc.

Analysis on the Interactive Meaning in the Verbal Language
Different from the previous two examples, this PSA consists of two images, one above and the other below. The verbal components of these two images is the same, which is on the lower right corner and upper right corner of the images respectively. The verbal message contained is this image include (1)"CIGARETTES SMOKE PEOPLE" with all the letters in uppercase and the sign and name of the institution who issued this advertisement which is (2)"Cancer Patient Aid Association". The verbal text (1) is a simple declarative sentence but with a logic quite unmatched with the rules of English grammar or daily conversation because in normal cases, this sentence should be "people smoke cigarettes", but not the other way round. With the subject and object switched around, the viewers may have a deeper impression on the image and be more easily influenced by this advertisement.

Analysis on the Visual-Verbal Inter-Semiotic Interaction
The visual and the verbal text of the advertisement cooperate to construct the interactive meaning of this image.
To begin with, in terms of visual contact, it is not sure whether this image is a "demand image" or an "offer image" because the presented participants' faces and bodies are "smoked up" by cigarettes; the close and far social distance between the two represented participants and the viewers embody different meanings; an oblique and low angle is adopted in terms of perspective; and this image is of high modality. As for the verbal text of the image, a declarative sentence is used to impress the viewers with unfamiliar logical relations. Therefore, the visual and the verbal text of the image integrate to address the topic issue -smoking as bad habits and to urge smokers to rethink their smoking habits to improve their health conditions.

Conclusions
Based on the previous analysis on three PSA examples in the field of wildlife protection, parenting, and health respectively, there are some major findings. Firstly, although there is some distinction in image representation, for example, in terms of visual contact, social distance, perspective and modality, the four aspects work together rather than separately to construct the interactive meaning. Their corresponding to each other is evidence to their cooperation. Secondly, the verbal text in the images tend to leave some space for contemplation to the readers. Through the way the producers organize the verbal text, the viewers may start asking themselves questions on the issue proposed, and may come up with their own new ideas on the current issues, but not reading somebody else's solutions to the problems without self-reflection. Thirdly, these images are generally of high modality to make them truthful and reliable. In terms of representation and contextualization, these images are all of high modality, with detailed environment description and exquisite background setting. Images of high modality turn out to be more truthful and can leave deeper impression on the readers, thus increasing the possibility of influencing their behaviors.
Needless to say, this study, like any other studies, has its limitations, which further research may refer to and make some improvements. Firstly, the example analysis part may feature subjectivity due to the author's personal ideas or the reason that some visual images are difficult to define and analyze in a precise manner. This could be a disadvantage of the study. For further research, the research methods for this kind of MDA may be improved and more other analytical methods shall be developed or used for MDA. Secondly, due to limitations on time and space, the data collected in the thesis may not suffice for the findings. For further research, more PSA examples shall be collected and analyzed for a more comprehensive study in the future. Thirdly, this study explore the static images only, while in this era of information technology, dynamic multimodal discourses such as videos, films, and television advertisements are prevailing. Therefore, for future research, dynamic multimodal discourse analysis shall be taken into consideration, studying how sound, action and other semiotic resources work together to construct interactive meaning. Lastly, due to the author's limited theoretical ability and rare literature, the findings in this study may need to be testified with the advances of academic research in the future.