CoMPoSition and FraMinG in ModErn and PoStModErn TIME Print adS: What haS ChanGEd and What thESE ChanGES MiGht MEan 1

This paper explores both possible changes regarding composition and framing in modern and postmodern tiME print ads as well as the meaning of these changes with respect to some of the main themes in the modern/postmodern debate. by comparing one-page ads from a 1929 and a 2009 tiME magazine issue, it identifies a shift from a balanced disconnected ( framing) top-bottom composition to two unbalanced but more connected ( framing) top-bottom compositions. Then, it attempts at interpreting this shift as a process of duplication of the modern compositional pattern, corroborating the position of social theorists who understand postmodernity as a phase within modernity itself.


Introduction
This paper explores what has changed in terms of composition and framing in modern and postmodern tiME print ads separated by a temporal gap of eighty years. More specifically, drawing on the concepts of composition, the informational values attributed to text and image according to their distribution on a given surface, and framing, the question whether text and image are represented as connected or disconnected elements on the multimodal page design (Kress & van leeuwen, 1996(Kress & van leeuwen, , 2006van leeuwen, 2005), it compares one-page ads from a 1929 and a 2009 tiME magazine issue in order to identify possible changes. The choice of working with tiME magazine ads is motivated by three main reasons. First, the fact that many scholars regard texts from the media, especially from print-media, such as newspapers, magazines and publicity materials, as one of the best sites for detecting cultural changes (Fairclough, 1995;Chouliaraki & Fairclough, 1999;Kress, 2003;iedema, 2003;bauman, 2007). Second, the fact that tiME magazine is a well-known US publication with wide circulation, being published without interruptions since 1923. Third, that regarding the modern/postmodern transition, the United States of america, due to its enormous economic growth after the Second World War, is one of the key countries in this context, if not the most relevant one (harvey, 1996). as for the issue dates, this choice has been motivated by the gradualness of the modern/ postmodern transition, which started in the 40's, during the postwar period, going up to the 70's/80's (Jameson, 2002;harvey, 1996). The dates selected, therefore, attempt to avoid the period of gradual transition. avoiding the transition period, however, does not mean the selection of "pure" samples of modern and postmodern ads. There is always a certain level of overlapping in the transition of social and cultural periods. besides the status of the modern/postmodern transition itself is still being intensively debated.
but that is not all. This paper also explores how these changes can be interpreted with respect to some of the main themes in the modern/postmodern debate. That is, after having identified possible changes regarding composition and framing, it attempts at establishing connections between these changes and three main themes: 1) the status of the modern/postmodern transition; 2) the way modernity and postmodernity are characterized; and 3) the centrality of the visual in both modernity and postmodernity. Congruently with Chouliaraki and Fairclough's (1999, p. 89) claim that "there is no such thing as a unified postmodern social theory" but only pertinent themes in the conceptualizations of the modern/ postmodern transition, the three themes explored in this paper are the ones that have been useful in the interpretation of the semiotic changes observed in the data analysis. a last word in this introduction concerns the organization of the paper. it is divided in seven sections. This introduction is the first and it presents the two-fold objective of the paper, its theoretical background and the data selected for analysis. The second section presents a quick view of each of the main modern/postmodern themes just introduced. The subsequent section brings an overview of all the possible categories regarding composition and framing, along with their particular meanings, and a brief note on the methods adopted in the analysis. This third section is illustrated with examples from the data to progressively familiarize the reader with the context of investigation. Then, the fourth and the fifth sections are dedicated to the analysis of composition and framing, respectively. The objective of these two analytical sections is to identify possible changes, which are subsequently interpreted in the sixth section. lastly, the seventh and last section draws the whole paper to an end by summing up its main points.
2. Three themes in the modern/postmodern debate regarding the first theme, and as pointed out before, there seems to be a lack of consensus on the status of the modern/postmodern transition. although the prefix "post" in the adjective postmodern, for example, seems to imply that modernity, or at least part of it, has been left behind, this is an idea upon which scholars do not seem to agree. Some say that what is regarded as modern is in some way already postmodern. For instance, in the arts, modern movements such as cubism, surrealism and conceptualism, in being an opposition to what comes before them, are already "post"-something. in this sense, postmodernity is not "modernism at its end but in the nascent state and this state is constant" (appignanesis & Garratt, 1995, p. 21). other scholars see modernity and postmodernity as being relatively distinct. They recognise that changes have taken place in the modern social historical condition, but they do not see a rupture deep enough to give birth to a new era. For them, postmodernity is, then, a term that serves to call attention to the discontinuities in the modern world and what society is living is an interim period between modernity and something not yet known (best & Kellner, 2001). Sharing this latter point of view, baudrillard (1983) offers an interpretation of postmodernity as a final stage of modernity, a stage of absolute power. according to the author (1983, p. 111), "there is no really radical difference between the two [modernity and postmodernity], only the scheme of control have become fantastically perfected. From a capitalist-productivist society to a neo-capitalist cybernetic order that aims at total control". in relation to the second theme, the way modernity and postmodernity are characterized, bauman's (2001,2007) metaphors, solid modernity and liquid modernity, seem to be considerably revealing and comprehensive. Solid modernity (modernity) stands for "an era of mass factories and mass armies", of "binding rules and conformity to rule" and also of "large volumes of spacious, heavy, stolid and immovable possessions" that are seen as a guarantee to a "well-anchored, durably protected and safe existence, immune to the future caprices of fate" (bauman, 2007, p. 29). Such modern emphasis on order, stability and security is easily explained by the previous scenario of infant mortality, death in childbirth, vulnerability to infectious diseases, social violence, little protection against natural disasters and short life expectancy. all this solidness, however, has also revealed a negative side effect. if reason, logic and order are efficient instruments in exercising some control, they also bring a great deal of uniformity, dullness and repetitiveness into human lives. in a clear contrast to solid modernity, liquid modernity (postmodernity) corresponds to openness to adjustment pressed by two main factors: the contemporary valuing of mobility, and the certainty of instability. These two factors may be illustrated with an example concerning contemporary employment perspectives. recent statistics point out that american youths who have completed secondary education are expected to change jobs at least 11 times in the course of their lives (bauman, 2001, p. 169). This obviously seems to be a lot when compared to the professional trajectory of modern workers who would start and finish their professional lives at the same company and, probably, performing the same roles.
lastly, with respect to the third theme, the centrality of the visual in both modernity and postmodernity, it is meant the increasing importance that the whole array of photos, drawings, cartoons, illustrations, diagrams and other visual resources that have become involved in the construction of different aspects of social life, have gained with the modern/postmodern transition. For instance, in the sphere of entertainment, many forms of leisure activity are now visually constructed (rose, 2001) and there seems to be a preference for them. Mirzoeff (as cited in rose, 2001, p. 8), for example, observes that at new york's Empire State building, the queues are longer for the virtual reality new york ride, than for the lifts to the observation platforms. and, in the field of literature, reflections on the consequences of a progressively visual culture, as for instance, in d. h. lawrence's short story The Blind Man (1920), in Carver's short story Cathedral (1983) and in Saramago's novel Blindness (1995), seem to have become a recurrent theme.
The three themes just presented are readdressed and/or further developed in the sixth section, where the semiotic changes are interpreted. in the next section, however, the focus is still on the theoretical background, more specifically, on the presentation of all possible categories regarding composition and framing.

Composition and framing
Composition and framing are semiotic systems initially conceptualized by Kress and van leeuwen in their Grammar of visual design (Gvd -1996(Gvd - , 2006, an extrapolation of halliday's Systemic Functional Grammar (SFG -1985(SFG - , 2004 to the domain of the visual semiotic mode, that have been further developed by van leeuwen in his work entitled Introducing Social Semiotics (2005) 2 . in general terms, these two systems are related to the way that different semiotic resources, for instance, image and text, are organized into a single and united message.
Composition, as explained before, has to do with informational values attributed to text and image according to their distribution on a given surface, which, in the present case, is the page of the print ad. in other words, left, right, top, bottom, centre and margins transfer their informational values to the images/texts placed on these pictorial areas. There are three possible distributions of text and image on the page of a print ad: 1) in a top-bottom composition; 2) in a left-right composition; and 3) in a centre-margin composition. in the case of the first, the top area is usually associated with "what might be" or the "ideal" and the bottom area is usually interpreted as "what is", the "real" (Kress & van leeuwen, 1996, p. 193). Kress and van leeuwen (1996) also state that, ideologically speaking, the ideal always plays a leading role, whereas the real plays a subservient role. The Studebaker ad (tlME 1929) in Figure 1 offers a clear example of a top-bottom composition. observe that the top area of the ad presents an image of a couple in front of a Studebaker, whereas the bottom area displays mainly a text. in this case, the image at the top plays the leading role for being associated with the ideal value, while the text at the bottom plays the subservient role for being associated with the real value. also observe that a dotted line has been added to Figure 1 in order to help visualize the top/ ideal and bottom/real contrast in this composition.  1929, p. 7. in the case of the second possible distribution, a left-right composition, the left side is taken as a familiar point of departure for the message, what is already "given". in contrast, the right is the side where "what is not yet known" is presented, the "new" piece of information (Kress & van leeuwen, 1996, p. 186, 187). The texaco ad (tiME 1929) in Figure 2 offers a clear example of a left-right composition. notice that, in this ad, the text on the left is presented as given, while the image and a small information box on the right are presented as new. again in Figure 2, a dotted line has been added to this composition to help visualize the contrast between the verbal given information (the text) and the new visual information (the image and the small information box). lastly, in the case of a centre-margin composition, the centre is considered to be the place for the most important piece of information, "the nucleus of information"; while information on the margins is seen as "contextualizing information" (Unsworth, 2001, p. 108). The boSE headphones ad (tiME 2009) in Figure 3 offers an example of a centre-margin composition. observe that, in this composition, the image of the headphones together with a heading ("Use them as a concert hall -or a sanctuary. ") work as the nucleus of information. The verbal and visual information on the margins (the pieces of text and the small images), on the other hand, take the role of contextualizing information. as before, a dotted circle has been added to the composition in Figure 3 to help visualize the opposition between the nucleus of information (the headphones together with a heading) and the contextualizing information (the pieces of text and the small images). in relation to framing, again, the question is whether text and image are represented as connected or disconnected elements on the multimodal page design. Framing is in fact a matter of degree and, regarding magazine ads, van leeuwen (2005, p. 6-14) has proposed an initial broad categorization under two main headings: 1) disconnection: when image and text are seen as belonging to different orders, and 2) connection: when image and text are seen as belonging to the same order. in the case of disconnection, the category is subdivided in three. The first one is segregation without overlap. That is, when text and image occupy entirely different territories. neither text nor image breaks through the opposite territory. The Studebaker ad (tiME 1929), previously presented in Figure 1, is an example of segregation without overlap. notice that, in this ad, there are no words in the top area (the image territory) and there are no images in the bottom area (the text territory). also the thick line around the image reinforces segregation between image and text. it is important to point out here that the date, the name of the magazine (tiME) and the page number that appear above a line at the top of the page, are not part of the Studebaker ad. all 1929 ads, with exception of those printed on the magazine covers, are always introduced within a frame, which includes these three pieces of information. This magazine frame, however, disappears in the 2009 data group.
Still in relation to disconnection, the second possibility is segregation with overlap. This occurs when text and image occupy different territories but somehow a part of an image or a piece of a text, for instance, a heading, invades the opposite territory. The texaco ad (tiME 2009), previously presented in Figure 2, is an example of segregation with overlap. observe that the right side of the ad, the image territory, is invaded not only by part of the main text but also by verbal information that comes in a box. Therefore, differently from the previous example ( Figure 1), framing in the texaco ad ( Figure 2) is porous.
The third and last possibility regarding disconnection is separation, which occurs when text and image are set apart by empty space, and this suggests that "they should be seen as similar in some respects and different in others" (van leeuwen, 2005, p. 13). The boSE headphone ad (tiME 2009), previously presented in Figure 3, is an example of separation. The image of the headphone is separated from the text only by empty space. There are no lines. in this example, image and text share the same white background. now, moving on to connection, this category is subdivided in two. one is pictorial integration, that is, when image is integrated in a predominately textual space, and the second is textual integration, when text is integrated in a predominately pictorial space, for instance, by being superimposed on an image. The drug Free america ad (tiME 2009) in Figure 4 is an example of pictorial integration. observe that, in this ad, the format of the page is basically that of a traditional running text preceded by a title (main heading). The image is integrated in the text by being inserted in the middle of the heading and three main columns. The auto-owners ad (tiME 2009) in Figure 5 is an example of textual integration. notice that, in this example, all textual elements are superimposed on the image of the award. notice, in special, the shadow under the piece of paper that contains the main text.  Summing up this brief description of composition and framing, table 1 brings all the possible categories and their respective meanings discussed so far. as for the methods adopted in the analysis, they consisted in identifying, classifying and quantifying all composition and framing types in the two data groups, the 1929 tiME ads and the 2009 tiME ads, according to the categories summarized in table 1.

Composition: from a single balanced top-bottom composition to two unbalanced top-bottom compositions
Starting with the examination of composition in the 1929 tiME ads, typical cases are those organized in a top-bottom composition (88%) in which image always appears in top position (100%), whether accompanied by a main heading or not. also, there seems to be a tendency (50%) towards a quite proportional division between top and bottom areas, that is, between image territory and text territory. Some examples of such proportional top-bottom composition with image in top position are presented in Figures 6 and 7. The first example is an ad for a variety of colourful plumbing fixtures produced by Standard Plumbing Fixtures. The second one is an ad for toothpaste, ipana toothpaste, which prevents gum bleeding. notice that in the two figures a dotted line has been added in order to help visualize the balanced top and bottom areas. in the case of Figure 6, a colourful image of a bathroom takes up the whole upper half of the page, being presented within a bluish and yellowish frame. in contrast, the bottom half of this page is predominantly textual. despite the presence of a small image of a sink (integrated in the text space), the company logo (bottom middle) and the addresses of four company branches (bottom sides), it is mainly a verbal text, preceded by a heading, that covers up most of the bottom section.
in the case of Figure 7, the scenario is not very different. a black and white image of a woman looking at herself in a mirror, along with a two-part heading, appears at the upper half of the page, while a four-block text comes at the bottom part. again, as in the case of Figure 6, the bottom half of Figure 7 also displays some other elements such as a small image of a toothpaste tube and a glass (bottom left) and a coupon (bottom right). This distribution of small images of products, coupons, company addresses and telephone numbers at the bottom area of ads has already been observed by van leeuwen (2005, p. 10, 200) and, according to this author, it is probably related to the more "down-to earth" aspect of the lower section of top-bottom compositions as previously discussed (cf. section 3).
Moving to the examination of composition in the 2009 tiME ads, typical cases are still those organized within a top-bottom composition (90%), in which image always takes top position (100%). The only noticeable difference here seems to be in terms of a lack of proportionality between top and bottom sections. Put differently, the previously discussed tendency of a balance between top and bottom areas in the 1929 data group is not present in the majority of the 2009 ads (78%). Figures 8 and 9 are examples of this lack of proportionality. Figure 8 is an ad for the nissan altima, a car model that undergoes 5,000 quality tests, including extreme temperature tests, in order to guarantee its durability. Figure 9 is an ad for the boSE acoustic Wave music system ii, a simplified one-piece music system that delivers high level sound performance. notice that a dotted line is again added to these two examples but this time to help visualize the unbalanced top-bottom compositions. in the nissan ad, Figure 8, it is possible to observe that the main image covers up much more than half of the page. The main text, on the other hand, comes at the bottom section occupying a narrow space and presenting a total of only 5 clauses 3 . in the boSE Music System ad, Figure 9, it is the main text that covers up most of the page space. This text presents a total of 33 clauses, within which the main image is integrated at the top right section. notice also that, similarly to the 1929 examples in Figures 6 and 7 8 and 9), the matter of proportionality may appear to present a slight difference between the two data groups. Such difference, however, has a very interesting implication that is now explored in more details.
Proportionality between top and bottom sections in the 1929 data group appears to have a direct influence on both the size of main images and the length of main texts displayed in these ads. This becomes quite evident when the number of clauses in the main texts of the 1929 ads is compared to that of the 2009 ads. Main texts in the 1929 data group range from a minimum of 4 clauses to a maximum of 25 clauses. to be more specific, half of the 1929 main texts present from 11 up to 20 clauses, a quarter of them present 4 to 9 clauses and the other quarter present 21 to 25 clauses. This quite limited range in the number of clauses per text is not found in the 2009 data group. Main texts in the 2009 ads may present a minimum of only one single clause up to a maximum of 99 clauses. More interesting, though, is that in the 2009 data group no ads present 11 to 20 clauses, the most common range in the 1929 data group.
based on the number of clauses per main text in the 2009 ads just discussed and also on the size of their main images, as described, for instance, in relation to Figures 8 and 9, it is possible to identify two different patterns in the 2009 tiME ads. one pattern, which accounts for the majority of the ads (60%) and which from now on is denominated the 2009 image Group, presents short main texts consisting of 1 up to 10 clauses and very large images. The nissan ad (Figure 8 Figures 11 and 12 stand for the two oppositional 2009 patterns, that is, for the 2009 image Group pattern and the 2009 text Group pattern, respectively. notice that the grey colour indicates "top image territory" and the black colour indicates "bottom text territory". also observe that the size of the ad page gets slightly smaller in the 2009 data group and that the magazine frame within which text and image are distributed in the 1929 data group disappears in the 2009 ads, as previously pointed out. to sum up, the analysis of the 1929 and the 2009 data groups in regards to composition has shown, on the one hand, a relatively stable scenario of top-bottom compositions with image always assuming top position. on the other hand, it has also shown an important change within this predominant organization, more specifically, lack of proportionality, which has generated two oppositional patterns in the 2009 data group.

Framing: from disconnection to some degree of connection
regarding framing, all 1929 ads (100%) present text and image as disconnected elements on the multimodal page. the two most common types of disconnection in this data group are segregation with overlap (56%) followed by separation (37.5%). Figures 13 and 14 are examples of disconnection in the 1929 data group. Segregation with overlap is present in Figure 13, an ad for old Colony, a trust company that started in the banana trade business and now offers services in the areas of banking and trading. Separation is present in Figure 14, an ad for SKF bearings, which have been successfully employed in a race boat called Miss america. observing the old Colony ad, Figure 13, despite some level of overlap, for instance, the small image of a funnel integrated in the bottom section of the text and two captions framing the main image, the two semiotic modes are still regarded as disconnected, mainly due to the thick black border at the base of the main image. in the SKF ad, Figure 14, disconnection becomes subtler but it is still present in this ad. in this case, image and text are not separated by a line, a border, or even by discontinuity of colour, but only by empty space. notice how each element in Figure 14 is surrounded by a halo of empty space that clearly delineates their territories.
although not as strongly as in the 1929 data group, disconnection is also predominant in the 2009 data group, accounting for 60% of the ads. The most common type of disconnection in these cases is segregation with overlap, which is present in 67% of them. Figures 15  and 16 are examples of segregation with overlap in the 2009 data group. The first example comes from the 2009 image Group and it is an ad for CiSCo, an american company that provides communications technology and services. in this ad, CiSCo announces that it will make a special appearance in the american television series "24", produced by Fox network. The second example comes from the 2009 text Group and it is an ad for rosetta Stone, a language learning software program. in the context of this ad, a teenager decides to use rosetta Stone in order to first learn italian and, then, impress an italian supermodel.  Figure 15) and in the rosetta Stone ad (Figure 16). From tiME, January 12, 2009, p. 3 ( Figure 15) and p. 46 (Figure 16).
in the CiSCo ad, Figure 15, it is mainly the edge of the image that forms a distinct boundary between the two semiotic modes. as it can be seen, the top section of the ad is basically image territory and the bottom section text territory. nevertheless, there is also some degree of overlap, with colourful dots forming an abstract shape at the bottom left side of the page and some textual information ("24 Just in time 2-night 4-hour premiere Sun Jan 11 Fox 8/7c") superimposed on the main image.
in the rosetta Stone ad, Figure 16, image territory is differentiated from text territory mainly by means of a thick colourful borderline right in the middle of the composition. both territories are then subdivided into smaller areas. at the top, the image territory is divided in two: 1) the black and white photo on the left side; and 2) the comment balloon coming out of the photo on the right side. at the bottom, the text territory is divided in three: 1) the yellow area right below the colourful borderline, which contains a list of languages; 2) the large white area in the middle that contains the main text; and 3) another yellow area with the rosetta Stone contacts and logo. as in the previous case, however, some level of overlap can be noticed in these subdivisions. For instance, in the bottom section, a small picture of the product and a notebook appear in the white area of the main text. also, at the top section, a heading in a small text box appears within the comment balloon claiming that the product offers "The fastest and easiest way to learn italian". differently from the 1929 data group, however, a significant tendency towards connection can also be noticed in the 2009 ads, accounting for the other 40% of these ads. it is interesting to observe that such tendency is exclusively translated into textual integration in the 2009 image Group and into pictorial integration in the 2009 text Group. Figure 17 is an example of textual integration in the 2009 image Group and Figure 18 is an example of pictorial integration in the 2009 text Group. The first figure shows an ad for hP 3G notebooks which allow people to work wherever they want to, "from around the corner to around the globe" as the main text in this ad affirms. The other figure shows the boSE Music System ad, previously presented in Figure 9. This ad announces a simplified one-piece music system that delivers high level sound performance. in the case of the hP ad, Figure 17, the whole ad page can be regarded as one big image. notice that the white text at the bottom of the page is integrated to this image by being superimposed on the reflection projected by the laptop. logos and some other institutional information are also superimposed on other darker areas of this image. in Figure 18, the situation is reversed. The format of the page is basically that of a traditional running text preceded by a title (main heading). here it is the image that is integrated at the top right section.
in brief, the results for framing analysis have shown a change from total disconnection in the 1929 ads (100%) to a more balanced situation between disconnection (60%) and connection (40%) in the 2009 ads. Possible interpretations of this change will be discussed in the next section. For the time being, however, the changes observed in respect to both composition and framing may be summarized as being from a balanced disconnected top-bottom composition to two unbalanced but more connected top-bottom compositions.

Interpreting the changes: duplication of the modern compositional pattern
This section starts with the matter of proportionality, which, as seen before, is quite a strong tendency in 1929 top-bottom compositions but not in the 2009 ones. Going back to the abstraction of the 1929 balanced top-bottom composition presented in Figure 10, it is possible to observe that proportionality between top and bottom areas seems to create some sort of horizontal symmetric equilibrium in this composition. That is, the top visual mass (image territory) is well balanced on top of the bottom visual mass (text territory). according to dondis (2000), symmetric equilibrium is a visual conception characterized by both absolute logic and simplicity that, occasionally, may run the risk of becoming inert and even dreary. notice that the symmetric top-bottom composition in Figure 10 could be easily achieved if one folded a page in half and left a constant margin. also notice that there seems to be no tension between the two pilled up rectangles since they have the same proportion.  Figures 11 and 12, respectively, it is possible to observe that horizontal symmetry is not present in them anymore. dondis (2000) explains that the Greeks regarded asymmetry as a precarious kind of equilibrium, but the author also points out that asymmetry is considered visually more stimulating and that it allows for variety. notice that in Figures 11 and  12, there seems to be some sort of tension between the disproportional masses, as if the larger ones pushed against the smaller ones. in other words, compared to Figure 10, the 2009 asymmetrical top-bottom compositions (Figures 11 and 12) seem to be more dynamic and, as a consequence, more exciting to the eye. relating these observations on the 1929 symmetric and the 2009 asymmetric top-bottom compositions to the way modernity and postmodernity are usually characterized, some interesting parallels seem to emerge. For instance, bauman's (2001,2007) previously introduced metaphors, solid modernity and liquid modernity (cf. introduction) appear to resonate well with the compositional values of inertness ( Figure 10) and dynamicity (Figures 11 and  12) just discussed. in solid modernity, similarly to the case of the 1929 symmetric compositions, logic and simplicity are regarded as important cognitive tools in the attempt to build a stable, predictable and, consequently manageable world (bauman, 2001). also, the result of such attempt may signify, in the long run, dull and oppressive routinization (eg. fordist factory). as for liquid modernity, the metaphor reflects mainly the openness to adjustment (bauman, 2001) that characterizes contemporary society and that is usually associated to ideas of both instability and mobility (cf. section 2). in a similar way, equilibrium in 2009 asymmetric compositions is regarded as uncertain or fragile, tending to a greater degree of dynamicity and allowing for more variety, which is now discussed.
Why are there two unbalanced 2009 top-bottom compositions instead of three or more? or, perhaps, instead of one like in the 1929 data group? a possible explanation may be in baudrillard 's (1983, p. 134-139) observation that "the matrix remains binary". That is, modern and postmodern systems have the number two as a common base. The only difference is that, in modernity, this binary matrix is expressed in terms of "a duel or open competitive struggle", the mechanism through which the system constantly challenges and transcends itself. in contrast, in postmodernity, the binary matrix is expressed in terms of "couples of simultaneous opposition", putting an end to the previous competitiveness and, more importantly, signalling the closure of the system in a process of duplication. baudrillard offers a very interesting discussion on Manhattan's architectural panorama as an illustration of these ideas. The author claims that, in modern times, all of Manhanttan's great buildings would be attacking each other in a kind of a "competitive verticality". however, with the construction of the World trade Center twin towers 4 between 1970 and 1971, competition has ceased to exist and the system has finally stabilized on a dual form: as high as they are, higher than all the others, the two towers signify nevertheless the end of verticality. They ignore the other buildings, they are not the same race, they no longer challenge them, nor compare themselves to them, they look one into the other as into a mirror and culminate in this prestige of similitude (baudrillard, 1983, p. 137). it is also important to point out that for baudrillard duplication is an essential strategy for any unitary system or monopoly that wishes to survive. The author makes clear that from "brands of soapsuds" to "peaceful existence", a monopoly can only achieve its final stage, that is, absolute power, when it is capable of "diffraction into various equivalents".
looking back at Figures 10, 11 and 12, it seems possible to make two valuable correlations between baudrillard's modern/ postmodern binary matrix and the three abstractions of top-bottom compositions shown in these figures. The first one is that Figures 11  and 12, the abstractions of the 2009 image Group and the 2009 text Group top-bottom compositions, seem to adjust well to the concept of "couples of simultaneous opposition"; not only because the two compositions come in a pair and stand in opposition to each other within the 2009 data group, but also because the types of asymmetry that they display are compatible instead of competitive: the 2009 image Group is top/image dominant and the 2009 text Group is bottom/text dominant. Perhaps, likewise the W.t.C twin towers in baudrillard's example, these two compositional patterns are also semiotic evidence of the closure of the system on a dual form and, consequently, of postmodernity as a final stage of modern monopoly, as often argued by scholars (cf. section 2). a second possible correlation has to do with the presence of competitiveness in the abstraction of the 1929 balanced top-bottom composition ( Figure 10). as discussed before, symmetric equilibrium in this composition appears to create a certain degree of steadiness and, possibly, dullness. nevertheless, examining this composition in more detail, it is interesting to notice that it somehow echoes the general layout of many different kinds of sport courts and fields designed for competitive events. in tennis, volleyball and basketball courts, as well as in football and soccer fields, for instance, a centralized line divides up a big rectangle into two proportional smaller rectangles, providing the same amount of ground for each of the individuals or teams engaged in "open competitive struggle" to use baudrillard's term. Therefore, despite the absence of tension in the 1929 balanced top-bottom composition, competitiveness appears to be indirectly related to the modern rationale behind it. Figure 19 brings the image of a multi sports court in order to help visualize the comparison. notice that, in spite of the different lines within the top and bottom sections, the half-court line is constant for all sports modalities. also notice that, as a spectator or a referee, one may have a horizontal perspective of the court/field, but as an active participant (a member of modern society) the perspective is always vertical as in the 1929 balanced top-bottom composition.
The discussion of baudrillard's modern/postmodern binary matrix may also be helpful in explaining a third aspect of the changes: the question of framing, more specifically, the shift from modern dominant disconnection to a certain degree of postmodern connection. Martinec and Salway (2005) have already established an analogy between the blurring of social boundaries in the postmodern scenario and the blurring of frames between image and text in the semiotic landscape. here this issue is just a little further developed based on baudrillard's terms. The point to be made is that sharp demarcations usually associated with modernity such as real versus unreal, inside versus outside, nature versus culture, high culture versus low culture, capitalism versus socialism and others (odih, 2007;Jameson, 2002;bauman, 2001) may be but a sign of its competitive binary matrix. in the modern pursuit for progress, the capacity to differentiate is vital. as in competitive sports ( Figure  19), in the end, one needs to tell, without a doubt, which the winner is. Thus, maintaining clear boundaries becomes an essential condition for the modern endeavour. now, in postmodernity, the system has already stabilized in a definite model "for the benefit of correlations" (baudrillard, 1983, p. 135, 136). no wonder, therefore, that framing becomes less sharp in the 2009 data group, allowing for some level of interrelationship between image and text.
Still on the interpretation of the changes, there are two final interrelated points. The first one regards the complete predominance of top-bottom composition in the data as a whole. as seen in the third section, this type of composition is present in 88% of the 1929 ads and in 90% of the 2009 ads. The second one has to do with the top position of images in the two data groups. as seen previously, regardless of the three types of top-bottom composition (Figures 10, 11 and 12), image territory has always been the top section (areas marked in grey), the place of the ideal element. one possible explanation for these unchanging characteristics is that they reflect the centrality of the visual to both modernity and postmodernity (cf. section 2). according to van leeuwen (2005, p. 204), the vertical plane upon which top-bottom compositions are organized is "the plane of spectacle, or of the façade of the building, the plane on which static categories are fixed and spatial order and structure created". also, according to the same author (2005, p. 205), elements presented in ideal position, in this case, images, are regarded as "the generalized essence of the information -(…) its ideologically most salient part". Such explanations go in line with our previous discussions on the increasing importance of the visual and with rose's (2001) observation that many different concepts permeating the modern/ postmodern debate share strong connotations to visuality such as scopic regime, society of spetacle (Guy debord), visual culture (Mirzoeff), simulacrum (baudrillard) and ocularcentrism (Jay).
to sum up, the interpretation of the changes observed seem to point at multiple correspondences between the different instances of semiotic change and some of the main themes in the modern/ postmodern debate. above all, however, these changes may be interpreted as evidence of a process of duplication (baudrillard, 1983). in other words, these changes seem to indicate that the single modern compositional pattern (top-bottom composition with image in top position) has expanded itself by means of generating two equivalents (two unbalanced top-botttom compositions with image in top leading position). in this process, breaking up with proportionality and intensifying connection have been the key strategies to move away from competitiveness towards compatibility.

Conclusion
This paper has started with a two-fold objective: 1) to explore possible changes regarding composition and framing in modern and postmodern tiME print ads; and 2) to explore the meaning of these changes with respect to some of the main themes in the modern/ postmodern debate. With respect to the first, it has identified a shift from a balanced disconnected top-bottom composition (1929 tiME ads) to two unbalanced but more connected top-bottom compositions (2009 tiME ads). and with respect to the second, it has interpreted this shift mainly as a process of duplication of the modern compositional pattern, corroborating the position of social theorists who understand postmodernity as phase within modernity itself. 1. in Kress andvan leeuwen's Gvd (1996, 2006), composition would be a broad system encompassing three subsystems: information value, framing, and salience. in this case, what is described here would refer to the subsystems information value and framing only. however, in van leeuwen's work (2005) composition is used as a synonym for information value. This paper adopts this last terminology because most of the interpretation regarding the changes are based on van leeuwen's work (2005).
2. Clause is the unit of analysis that halliday (1985,2004) establishes for his SFG.