Reference and Multimodality: analysis of the construction of reference in advertisements for vehicles

The objective of this article is to analyze the construction of the referent “vehicle” in advertisements, proposing an interface between Referencing and Multimodality. For this, it follows the sociocognitive-discursive approach of Referencing (CAVALCANTE, 2011, 2012, 2015; KOCH, 2004, 2015) and uses, as a foundation on Multimodality, the Kress and van Leeuwen’s (2006) Grammar of Visual Design (GVD). It analyzes, through the compositional metafunction, formed by the subcategories value of information, salience and framing, the referential construction in four advertisements for vehicles. Considering the analyzes carried out, it concludes that the categories salience and framing occur in a complementary way in the construction of the Referent, becoming, in a way, redundant, because what is arranged in the image as a disconnected element and therefore functioning independently, coincides with what is evidenced, placed in a prominent position, most of the time, centralized in the image. Specifically, regarding the analyzed discursive genre, it concludes that the referent “vehicle” is the nucleus of visual information, confirmed by the deletion effect suffered by the elements in the background. The results obtained indicate that the visual composition maintains a direct relation with the construction of the Referent.

Research on multimodal discursive genres gained momentum from the Grammar of Visual Design (hereafter GDV), proposed by Kress and van Leeuwen (2006) 1 . Based on the systemic-functional Linguistics of Halliday (1987), the authors of the GDV propose analytical categories that allow the reading of images through a visual syntax, since, according to them, everything that is expressed, linguistically, through words and linguistic structures, can be expressed, imaginatively, through the use of colors and different compositional structures.
In this sense, the objective of this article is to analyze the construction of the vehicle referent in advertisements, collected on websites of two major automotive automakers, namely Fiat and Volkswagen, for which we propose an interface between Referencing and Multimodality. For this, we will follow the sociocognitive-discursive approach of Referencing (CAVALCANTE, 2011(CAVALCANTE, , 2012(CAVALCANTE, , 2015KOCH, 2004KOCH, , 2015 for the analysis of the referent and will use, as a foundation on Multimodality, GDV, proposed by Kress and van Leeuwen, 2006). More specifically, with respect to the compositional meta-function, formed by the subcategories information value (Given-New, Ideal-Real and Centre-Margin), salience and framing, to analyze the construction of the referent in four vehicle advertisements, which will represent what we detected in the research corpus in general, for which we analyzed 85 ads.
This article is organized in four sections: in the first, it presents the current perspective of the studies about the Referencing; in the second, we present the contribution of the GDV to the visual syntax studies; in the third, we explained the methodology used; and, in the fourth, we deal with data analysis. Finally, we present our final considerations.

REFERENCING: AN IMBRICATION BETWEEN SOCIETY, COGNITION AND DISCOURSE
In this section, we present the current approach on Referencing, which includes the studies developed by Textual Linguistics (LT), adopted in this work. According to Cavalcante (2015), LT relies on the conception of an intricate cognition with the social environment and with the discursive practices to which they support, which it defends as sociocognitivism, therefore breaking with a representational view of language. By conceiving language as resulting from the interrelationship between social interaction, cognition and discursive practices, we consider this sociocognitive-discursive approach.
The conception of Referencing, from a sociocognitive-discursive perspective (CAVALCANTE, 2011), used in this work, is based on the thinking of Mondada and Dubois (2003), who conceive the referent 2 as an object of discourse, that is, as a created object not by discourse, defining it as an inherently unstable cognitive-discursive category. According to this view, the referent is not given a priori, but is constructed collaboratively among the interlocutors during the interaction.
In this sense, Marcuschi (2008, p. 142) explains that: The reference, as treated by Mondada and Dubois (1995), is a process carried out in the discourse and resulting from the construction of the referent, so that the notion of reference becomes a different use from that which is attributed to it in semantic literature in general.
Mentioning is not an activity of "tagging" one extensionally preexisting designated world, but a discursive activity (essentially creative), so that the referrings become objects-of-speech.
Therefore, the Referencing can only be understood as a discursive activity, since the discourse constructs what it remotes, while at the same time it is tributary of that construction (KOCH, 2004). Besides that, Koch (2004) adds that every discourse constructs a representation that operates as a shared memory, in which its successive stages of representation are responsible for the selections -referential expressions -made by the interlocutors. The way in which the interlocutors make their choices to refer to the discourseobjects, that is, to the forms of reference "are choices of the subject in interaction with other subjects, in function of a want-to-say" (KOCH; ELIAS, 2012, p. 124). This assumption is essential for referential construction as a complex activity resulting from interaction between interlocutors.
This view is also shared by Cavalcante (2012), who conceives Referencing as a negotiation between interlocutors, as a negotiated, cooperative and intersubjective process, which aims to meet the needs of interaction; it is a sociocognitive phenomenon. The initial view of reference as being the relation between thing and object, therefore it does not contemplate the pragmatic and discursive aspects.
Thus, we consider that the sociocognitive-discursive approach allows us to understand Referencing in all its complexity, providing better conditions for the comprehension and production of the meaning of the text, corroborating the notion that such meaning is constructed in text-subject interaction (KOCH, 2015). In this sense, we emphasize that the intrinsic dynamicity of the Referencing phenomenon is also due to the way in which the Referent is introduced and recategorized, since it can undergo modifications in the text.
When we speak of Referencing, we consider it opportune to clarify the notion of Referent as an abstract entity, built in and by discourse through the interaction between interlocutors, in which we guide our discussion. For Cavalcante (2011, p.15), the Referents are Entities that we construct mentally when we enunciate a text. They are abstract realities, therefore immaterial. Referents are not meanings, although it is not possible to speak of reference without resorting to traces of signification, which inform us of what we are dealing with, what it is for, when we employ, and so on. Referents are not forms, although, in general, they are carried out by reference expressions Therefore, the Referent is an object of discourse and not an entity of the extralinguistic world, whose construction depends on a joint action among the interlocutors, since it is "in the interaction, mediated by the other, and in the integration of our language practices with our sociocultural experiences that we construct a representation -always unstable -of these entities that are referred to as referents" (CAVALCANTE, 2011, p. 15-16). Therefore, it is through language that the phenomenon of Referencing is realized. In this way, Cavalcante (2012) affirms that the process of construction of the Referents implies that, "in the end, the role of language is not to faithfully express a ready and finished reality, but, rather, to construct, through language, a version, an elaboration of events occurred, known, experienced" (CAVALCANTE, 2012, p. 105).
In this construction, the Referent can go through modifications, transformations, that is, it can be recategorized, aiming at attending to different discursive functions. Regarding the recategorization process, Lima (2009) states that this is a process of a cognitive-referential nature, which derives from the categorization in the reference activity. This presupposition has the following consequences: i) recategorization cannot always be rebuilt directly at the textual-discursive level, not configuring only by the remission or resumption of lexical items; (ii) in cases where (i) recategorization must be (re) constructed by evoking elements rooted in a cognitive level, but always signaled by linguistic cues, in order to avoid interpretive extrapolations; iii) as a result of (ii), recategorization may have different degrees of explicitness and necessarily imply inferential processes (LIMA, 2009, p. 57).
Thus, recategorization is not limited to the textual-discursive level, characterized by semantic and pragmatic aspects, in order to meet the communicative purposes of the text producer. This conception, therefore, is aligned with the sociocognitive-discursive approach of the phenomenon of Referencing.
Thus, the Referent, as a mental representation of an entity established in the text, can manifest formally through referential expressions, which are, in most cases, linguistic structures of a substantive nature; by means of images, for "as far as the reference, the visual resources of a text can perform functions similar to the linguistic resources" (CAVALCANTE; CUSTÓDIO FILHO; BRITO, 2014, p. 43); but it may not be formally explicit in the text, since it is a discursive construction, and recategorized to effect the process of Referencing.
These are, therefore, the reference processes 3 most commonly addressed in the studies on Referencing: referential introduction and anaphora. The first, as the name reveals, is the moment when the Referent appears for the first time in the text. The second is configured by everything that retakes the Referent (CAVALCANTE; CUSTÓDIO FILHO; BRITO, 2014).
The two processes can be seen in the excerpt from text by Carlos Drummond de Andrade. In the example, the first expression flower, present in the title, constitutes a reference introduction (in italics), while the other expressions that refer to the discursive object "flower" constitute anaphora (in bold). Let us also observe that each referential expression, in a certain way, takes up the Referent "flower", adding other senses to it, thus, recategorizing it. Let us look at the passage in question: Flower steal I stole a flower from that garden. The porter of the building was dozing, and I stole the flower. I brought it home and put it in the glass with water. I soon felt she was not happy ... (ANDRADE, 1985, p. 80).
In this text, in which verbal language predominates, both referential introduction and anaphora are easily identifiable. However, it is not always possible to identify them as easily in multimodal texts as verbal-imagery, since in these texts, as Silva (2013) states, referents can be introduced both by the image and by the reference expression.
Given the complexity and dynamism inherent in the phenomenon of Referencing, in the next section, we will discuss Multimodality in order to better understand the nature of multimodal discursive genres, and then analyze the construction of the referent in a genre constituted by the imbrication of the verbal language and imaging language.

MULTIMODALITY AND VISUAL SYNTAX
In this section, we briefly present the elements that characterize the Grammar of Visual Design (GDV) and its metafunctions, in order to locate the multimodal category that we will use in the analysis of vehicle advertisements.
The GDV arose in order to meet the needs of the contemporary world, full of multisemiotic texts, which lacked analytical categories that could facilitate their reading and comprehension, given that, the primacy of studies focusing on verbal language leaves little room to the study of multimodal texts.
We emphasize that GDV is part of Social Semiotics, a theory dedicated to the study of representation and human communication, according to its authors. These are based on Systemic -Functional Linguistics (LSF), proposed by Halliday, which establishes three metafunctions: the ideational, the interpersonal and the textual. From these metafunctions, Kress and van Leeuwen (2006) establish the ideational, interactive and compositional metafunctions.
The GDV authors explain the metafunctions proposed by Halliday in the following way: ideational metafunction describes the way in which different semiotic modes represent aspects of the world as experienced by humans, "semiotic modes offer an array of choices, of different ways in which objects, and their relations to other objects and to processes, can be represented" (KRESS; VAN LEUWWEN, 2006, p. 42). Therefore, the ideational metafunction is linked to the representation function of the world outside of individuals. The interpersonal metafunction explains the capacity that any semiotic mode must have to represent social relations between the producer, the spectator and the represented object, and the textual metafunction clarifies the meanings resulting from different compositional arrangements of the text. For the authors, the textual metafunction describes the ability of the different semiotic modes to "form complexes texts of signs internally coherent with each other and externally coherent with the context in and for which they were produced" (KRESS; VAN LEUWWEN, 2006, p. 43).
In order to better understand these metafunctions, we can observe the chart proposed by Fernandes and Almeida (2008), in which they summarize the metafunctions proposed by the Systemic-Functional Linguistics (Halliday) and by Multimodality (Kress and van Leeuwen), relating them: Among the three metafunctions, for this work, we select as analytical category the compositional, which highlights the way the visual elements are presented in the image. The choice for this metafunction is justified by the fact that composition, according to Kress and van Leeuwen (2006, p. 177), relates the representational and interactive meanings of the image through three interrelated systems: information value, salience and framing. As the name itself reveals, this metafunction aims to analyze the composition of the image from the distribution of its elements. Let's look at each element: (1) Information value. The placement of the elements (participants and syntagmata that relate them to each other and to the viewer) endows them with specific informational values attached to the various "zones" of the image: left and right, top and bottom, center and margin; (2) Salience. The elements (participants as well as representational and interactive syntagmata) are made to attract the viewer's attention to different degrees, as realized by such factors as placement in the foreground and background, relative size, contrasts in tonal (or color) value, differences in sharpness etc.;  VAN LEEUWEN, 2000), where visual resources construct "the nature of the viewer's relationships and what is seen." TEXTUAL COMPOSITIONAL responsible for the structure and format of the text, it is performed in the compositional function in the proposition for image analysis of Kress and van Leeuwen, and refers to the meanings obtained through the "distribution of information value or relative emphasis between the elements of the image.
(3) Framing. The presence or absence of framing devices (realized by elements which create dividing lines, or by actual frame lines) disconnects or connects elements of the image, signifying that they belong or do not belong together in some sense. Thus, with regard to the information value, the chart is organized from: • a horizontal line that divides it into the upper and lower part. In the first, we find the information considered as ideal and, in the second, the information considered as real; • a vertical line dividing it into the right side and the left side. In the first, we find the information considered as given or already known and, in the second, the information considered as new; and • a more central area, in which the most relevant information is found, and a more marginal, peripheral area in which complementary information is found. Therefore, the triptych consists of the pairs: Given / New, Ideal / Real and Centre / Margin, which are related in the construction of the meanings of the text. (See triptych later) These elements are arranged in the following chart, in which one can observe the triptych, composing the dimensions of the visual space: Chart 2 -Dimensions of visual space Source: Kress e van Leeuwen (2006, p. 197).
However, we note that in many texts there is no perfect symmetry, as triptych seems to suggest (see Chart 2). The texts enjoy flexibility at the discretion of the authors' intentionality, that is, the lines ideally delimit the information. In actual production, the center, for example, can move further to the right or left, up or down. This, however, does not misconceive information status or positioning. The announcement below, Figure 1, for example, regarding the information value -center/margin, there is no central region highlighted.
Source: http://bit.ly/2FxF6Ee Figure 1 -Value of the information cent / margins As for the criterion concerning the salience, that is, the way certain elements are highlighted in the image, this is, in a way, intuitive, because as the authors state, "The viewers of spatial compositions are intuitively able to judge the 'weight' of the various elements of a composition, and the greater the weight of an element, the greater its salience" (KRESS; VAN LEEUWEN, 2006, p. 202). The authors point out that this salience is not objectively measurable because it results from a complex interaction, from a complex negotiation relationship between a number of factors: size, focus sharpness, tonal contrast, color contrast, visual and factor field positioning cultural practices. It should be noted that the composition of a picture or a page also involves different degrees of salience to its elements. Regardless of where they are placed, salience can create a hierarchy of importance among the elements, selecting some as more important, worthier of attention than others. The Given may be more salient than the New, for instance, or the New more salient than the Given, or both may be equally salient. And the same applies to Ideal and Real and to Centre and Margin (KRESS; VAN LEEUWEN, 2006, p. 201).
The salience, then, becomes an important resource for the discursive genres as the advertisements. Thus, Sousa (2016, p. 100), based on Kress and van Leeuwen (2006), when analyzing this criterion, states that "the salience is a way of drawing the attention of the reader to a segment of the text in detriment of other passages". The author, in analyzing the advertisement of a pizzeria, Figure 2, observes that the salience is accomplished through resources such as the colors, size and shape (round and full) of the pizza image; and other resources, such as bold, capital letters at the beginning of each word, before the first sentence, font size and varying colors (stronger, softer, etc.). Source: Sousa (2016, p. 103).

Figure 2 -Information Value salience
Finally, with regard to framing, we have that the connection or disconnection between the elements of the composition contributes to be seen as isolated units or not. According to Kress and van Leeuwen (2006), the more elements of spatial composition are connected, the more they are presented as a single unit of information.
According to Sales (2017), salience and framing criteria are related, since "when one element of the text is disconnected from the others, there will be a greater salience, then a framework, because the element will be highlighted, in a frame" (p. 49). Moreover, in her analyzes, the author states that the relationship of framing and salience are the ones that most indicate referential introductions "due to the fact that the disposition of the images focus on what would most interest the reader", so that the salience becomes "an aspect that generates a focus for the referent, being able to present it as a reference introduction as well as to emphasize the relation of figure/background" (p. 79).

Figure 3 -Information Value framing
In the light of the above, we point out, first of all, that the combination of these criteria -information value, salience and framing -is flexible, because in some images one of them may predominate more than others and, secondly, they help us to understand the visual syntax that allows the construction of different senses of the text.
Thus, we see that the visual syntax provides different readings and, consequently, different constructions of meaning of the text, reason why we recognize as legitimate the convergent relation between the Referencing and the Multimodality, both acting in a complementary way in the investigation of phenomena of interest of Textual Linguistics, as in this case in which we deal with the construction of the referent.

THE CONSTRUCTION OF THE REFERENT VEHICLE IN ADVERTISEMENTS
Given the diversity of advertisements for vehicles circulating in different media, we selected 04 advertisements of different brands of vehicles, published on Brazilian sites of the automakers, taken from the corpus of our research, to analyze in this article. The advertisements will be analyzed from the category called visual composition, in which we have (i) the information value -given and new, ideal and real, center and margin; (ii) the salience and (iii) the framing, as to its participation in the construction of the vehicle reference.
In advertisement 01, besides the image, we find the verbal text: "GOLF. IT IS NOT JUST A CAR. IT'S THE GOLF " Advertisement 1 Source: http://bit.ly/2ZO6d5C Notice, in Chart 3, the description we make of ad 1.
Chart 3 -Descriptive analysis: Advertisement 1 Source: Prepared by the authors.
The construction of the referent vehicle in this advertisement occurs, therefore, by the interrelation between the meanings constructed through the metafunctions. Thus, for example, the deletion effect (fuzzy features) of the background of the image, in order to highlight the vehicle, can be understood, among other possibilities, as being intended to make the reader see only the vehicle and not distract with elements of the landscape, as can also refer to engine power, suggesting that this car is ideal for long journeys by being fast and safe, as well as comfortable. It is possible to state that the image of the vehicle itself COMPOSITIONAL METAFUNCTION

Information Value Salience Framing
It predominates the center/margin pair, in which the vehicle is in the central region, composing a landscape that refers to a road that crosses mountains, but that can hardly be seen due to the speed effect created in the image. Within the Ideal/Real pair, we highlight the information that this is the ideal vehicle for those who like to "get on the road".
The salience of the vehicle is made by erasing the background, formed by the landscape, to accentuate the visibility and sharpness of the car in motion. The elements such as color, focus and vehicle brightness contrast with the opacity (diffuse images due to the car's speed) of the background, making it stand out.
The image of the vehicle is disconnected from the rest of the landscape, so its frame is considered weak, because, as Sales (2017, p. 78) reminds us, "the more prominence, salience, is given to an element of the text, less framing will be in textual composition." However, in this case, the disconnection is important to make sense (message, which the advertiser wants to pass: fast, powerful, safe car).
provides other information, such as being a beautiful car, with a modern, bold and sporty design. Thus, if we follow the visual syntax constructed by the advertisement, we will develop a reading that starts from the upper left and goes to the center, so that verbal language provides the "ending" of the construction of the intended meaning, since in expressing "GOLF. IT IS NOT JUST A CAR. IS THE GOLF", we can infer the following reading: 1) the referent is introduced through the referential expression Golf vehicle, thus an introduction already presenting the vehicle, naming it. The strategy of presenting it by name suggests a way to bring the reader, a possible buyer, closer to him because it creates an already familiar context of intimacy. 2) Then, the referent receives a first recategorization 4 -"It's not just a car," suggesting it has more features and attributes than regular vehicles, so it's different. 3) Finally, the expression "Is the Golf ", when winning the determinant, the advertiser seeks to emphasize that this vehicle, being widely known in the market, transmits confidence and credibility to the reader. If, before, the reading allowed us to interpret that the Golf is not only a car, now, we have that the Golf is not any car. The reader, therefore, is led to infer that the vehicle has characteristics that make it bearer of excellent qualities in front of the others.
Such choices in the composition of the advertising genre suggest special care of the advertiser with their intended audience, usually made up of clients from a younger age group.
Let us look at Advertisement 2, below:
The convergence of these elements suggests a reading that begins with verbal language, that is, with the text "NOVO JETTA. SEDAN'S DISCOVERED SPORT". This route leads us to understand the expression "New Jetta" as the referential introduction. Thus, the reference Jetta is introduced as being a new vehicle, implying that it is a better model than its predecessor. Then, the referent is recategorized, as it gains the attribute of being a "sport vehicle disguised as sedan". In this way, the reader is led to trigger their knowledge of the world and infer that a sedan car is characterized by being large, sophisticated and comfortable, and therefore generally more expensive. On the other hand, sports cars are considered modern and, commonly, sought after by a younger audience.
We emphasize that the typography used in the expression "NEW JETTA", in our opinion, confirms our reading that this is the resource that constitutes the introduction of the referent.
This information licenses the construction of the referent as an affordable, simple, but at the same time modern, sophisticated and luxurious vehicle. At this point, we consider that the image of the elegantly dressed woman located at the margin somehow contributes to this construction, since the image refers to a young person who dresses in a simple but elegant way, in line with the characteristics of the vehicle.
Thus, the choices made by the advertiser in the production of this advertisement reveal that the overlap between verbal and imagery languages produce interesting effects of meaning in order to persuade his client.
Let us go to Advertisement 3:

Information Value Salience Framing
We have the verbal language occupying the Ideal information space (upper part) and Given (left side); and the image at the center, showing the vehicle as the most important referent in the advertisement.
At the edges, we see a building and a woman dressed in an elegant way, setting elements less salient.
The vehicle is in the center, it is presented with a gleaming gray color, contrasting with the scenery color around it. The sporty rims that attract the attention of the sporting enthusiast are particularly noteworthy, as well as the knowledge that vehicle lovers appreciate that this type of rim adds value to the vehicle.
Although other elements appearbuilding and woman -the vehicle does not connect with any of them. Therefore, those elements appear distanced and without much detail, while the vehicle appears in the foreground and with clarity. Here again, the framing performed by the disconnect is responsible for highlighting the vehicle referent.
In this case, it seems to us, the beginning of the reading is authorized by the image, leading the reader to seek more information in the verbal text, thus the image is configured as referential introduction and the verbal language is configured as recategorization. Thereby, the name of the vehicle -Punto -recategorizes the vehicle referent, since when you name it, you specify it. Then, the referent is once again recategorized by means of the expression "MORE SPORTS WITH PERSONALIZATION DIFFERENTIATED", homologating the construction of a reference vehicle Punto as being sporty and differentiated, suggesting attending the taste of the possible buyer.

Information Value Salience Framing
As in previous announcements, the image of the vehicle appears in the center, while the name is subtly displayed at the bottom left, indicating that it is a given information, that is, the vehicle is already known; at the same time, being at the bottom, as being real, indicates something accessible to the reader.
The luminous lines that seem to come from above, simulating something like a spacecraft, are directed to the vehicle by highlighting it. The black and white colors harmonize in a game of contrasts that enhance the brilliance of the car.
The lines and the circle place the vehicle in the center of the image, making clear the intention to offer this product. Unlike the other ads in which the vehicle is disconnected, here it appears as totally connected with the scene, mainly, by the way the lines follow to the circle in which the vehicle is, also by the contrast of the colors black and white, used both in the scenery and in the vehicle.
The advertisement reveals a composition dominated by the harmony of the black and white colors that make it up, so the imagery language stands out, while the verbal language appears in small letters, almost illegible, except for the name of the vehicle "Punto". Please note that, on the website where the ad is located, the reading is compromised due to the small size of the font used and it needs to be enlarged to make reading easier. We consider this strategy as a resource that induces the reader to dwell on the vehicle's image, since it is produced to persuade the prospective buyer. However, anyone who feels instigated to read the verbal text will probably do so by seeking more arguments that convince him to make the decision to buy it.
Thus, the ad carries the following verbal text, which we prefer to reproduce in the same arrangement in which it appears:

Punto
More sportsmanship with differentiated customization. Punto has a modern, sporting car look. The design reinforces the strong personality and accentuated style that have always been the essence of the car. Between the versions Attractive 1.4, Essence 1.6, Essence Dualogic 1.6, Blackmotion Dualogic 1.8, you will find a car as authentic as you. If you're picky about design, you'll love Punto.
The typography used, as well as the difference between the sizes and the layout of the source, reveals a sequence of information. In this way, the most important information -the name of the vehicle -appears bigger and highlighted, followed by a second important information that seeks to persuade the prospective buyer through the argument that the car is sporting and differentiated. Finally, there is a verbal text on design, which presents other models of this line of vehicles.
We consider that by using these strategies in the composition of this discursive genre, the advertiser intends to reach a specific audience, that is, a younger audience interested in sports cars.
A similar situation can be seen in Advertisement 4, since the highlight is given to the vehicle Palio, through the image in the center, situating it as the nucleus of the information, thus as the referent. In the same way, we present the transcribed verbal text to facilitate reading, since it becomes essential for the construction of the intended meanings in the text. Let us move on to the analysis: Due to the very small size of the information contained in the lower left corner of the advertisement, we consider it relevant to transcribe them, due to its importance in the construction and recategorization of the referent:

Palio
All comfort and practicality you deserve. If Palio was already excellent, now Palio 2017 is even better.
Combining an innovative design with the touch of the Italian personality, Palio arrives with the most sophisticated interior.
Chart 6 -Descriptive analysis: Advertisement 4 Source: Prepared by the authors.

Information Value Salience Framing
In this advertisement, we have the center margin pair protruding up, so that the car is in evidence in relation to the bottom.
The name Palio is a known information and places the vehicle close and accessible, as it is in the lower left corner.
The white color of the car contrasts with the dark background of the image, making it stand out. The car stands alone, while the buildings in the background form a single block. The glow radiating from the headlights suggests attracting the reader's gaze.
The vehicle is positioned forward with no connection to the rest of the image. There is a visible line separating it from the city image in the background. The buildings seem to form a single block while the vehicle appears isolated, advertiser's strategy to highlight the vehicle.
The meanings built from the arrangement of the verbal-pictorial elements license the construction of the reference vehicle Palio as being a modern and sophisticated automobile. If we start reading through the imagery language, this will configure the introduction of the referent, and the verbal language, the recategorization, which is given by the addition of information (innovative design with the touch of Italian personality, more sophisticated). In addition, verbal language brings the adjectives -excellent, better, innovative, Italian and sophisticated -creating semantic relationships that recategorize the referent, in order to persuade the prospective buyer.
In this advertisement, the choices made by the advertiser suggest to the intended audience a vehicle that is both practical, modern, comfortable and accessible, therefore ideal for the life of big cities.
Considering the above, the analysis of these advertisements suggests that the visual composition, by relating the categories information value, salience and framing, contribute to the introduction of the Referent "vehicle". Regarding the information value, it was possible to notice that the use of the Center-Margin relationship predominates, in which the vehicle occupies the central region of the advertisement, a procedure carried out by means of a deletion effect of the other elements or positioning them peripherally.
As for the Ideal-Real pair, we notice that, in advertisements 01 and 02, the verbal language is somewhat relevant, since it is positioned in the upper left quadrant, so that the Referent is presented as an idealized product. We consider that this strategy aims to value the vehicle, offering it as an object of desire to a possible buyer. On the other hand, in the advertisements 03 and 04, the focus is on the image, since the verbal language is almost imperceptible, except for the name of the vehicle. Thus, the Referent "vehicle" can be noticed by the reader (possible customer) as a Real information, suggesting that the product, object of desire, is already available, within reach of the customer.
In advertisements 01 and 02, the verbal language is noticed occupying the Given information space, but there is no symmetry in the image in relation to the New information. This fact reinforces that the triptych is not something rigid; on the contrary, it is flexible, in order to meet the strategies of persuasion of the text producer. In this sense, we consider that the image of the vehicle positioned at the Center leads us to take it as the New information.
Advertisements 03 and 04 focus on the information Given and already established as something real, existing. This choice suggests the presentation of the Referent "vehicle" as a product already known, possibly associating it with the knowledge that these automobiles, as well as the quality of them, are known by the possible client.
In short, the different possibilities of using the dimensions of the visual space, that is, the division of this on the left side-right side, upper part-lower part and center-margins, in verbal-pictorial texts, such as advertisements, binding between the verbal and imaginary languages, allow the construction of the Referent from different effects of meanings that seek to persuade potential buyers.

FINAL CONSIDERATIONS
In the face of the analysis, we notice that the categories salience and framing occur in a complementary way, becoming somewhat redundant, because what is arranged in the image as a disconnected element and therefore functioning more independently coincides with the evidenced one, placed in a prominent position, most of the time, in the center of the image. This observation corroborates the findings made by Sales (2017), when analyzing the discursive genre charge.
Specifically, as regards the analyzed discursive genre, we perceive that the Referent vehicle is the nucleus of visual information, confirmed by the deletion effect suffered by the elements, in the background, as we have shown.
In this sample of advertisements, the analysis reveals that the Referent predominates in the center of the triptych, that is, in the central region of the composition, placing the vehicle always in focus. In advertisements 01 and 02, the verbal language appears in the upper left corner, therefore, as being Given information, already known by the viewer and, at the same time, as Ideal, so we understand that the intention of the advertiser is to emphasize the "promise of the product "as something already given, that is, what "could be" is already possible to enjoy in the advertised vehicle, the viewer does not need to desire a perfect car, since this, in fact, already exists.
Typographic features, such as font type and size, colors, glows, contrasts and visual effects, contribute greatly to the construction of the senses, as they reveal the most important information, pointing to the elements that present themselves as a referential introduction and those in which the process of construction and recategorization is concentrated.
In short, we consider that the categories proposed by the GDV contribute to the understanding of the phenomenon of the Referencing, allowing relevant analyzes on the way the referents are constructed and recategorized in multimodal discursive genres.