Beyond Linearity: Holistic, Multidirectional, Multilinear and Translinear Reading in Comics

This article presents a discussion about some of the main theoretical approaches of the assemblage of panels on the page and the double page, arguing that the correspondences between the images on the page are not fundamentally linear. On the contrary, comics foster readings that can be holistic, multidirectional and multilinear. Moreover, the correspondences between images through the pages of the whole book are also relevant. The book form, often disregarded by the critics, actually plays a very important role in the way comics are read and favour a reading that is translinear. Finally, some works are even breaking with the codex as a format, suggesting new ways of reading and engaging with the comics form. In order to illustrate how all these reading protocols work, this article offers a selection of examples of graphic narratives from different periods, from Winsor McCay to Richard McGuire.

productive (the story and the layout are interdependent and work together, often destabilizing the conventional reading order) (2003: 51-75). However, some critics have questioned this classification. Thierry Groensteen, for instance, has re-evaluated the conventional use of the layout and has suggested another typology focused on the general opposition of discrete units (predominantly sequential, dominated by the story) (2007: 97-102) and, in a later work, has extended the typology to levels of regularity and expanded the rhetoric uses (2013: 43-47). This article tries to expand on these typologies by focusing, not only on the page, but on the book's form.
While it is easy to assume that comics that use more regular page layouts also tend to make use of a more lineal and conventional reading, this is not always the case, as this analysis will show. There are, however, clear connections between the page layouts and the way readers processes them when they read a comic. The traditional Z-pattern, which follows a left-right, top-bottom way of reading, 1 has been clearly inherited from the process of reading text in book format and has been highly influential when reading comics. However, this pattern is frequently altered, mainly when the grid is broken in page layouts, a technique that Neil Cohn calls blockage (a column with two equal frames followed by a column with a frame occupying the same vertical space of the previous two) and overlap (overlapping frames) (Cohn 2013: 93-95). Cohn affirms that there is a principle of assemblage, according to which the reader tries to create units based on the distance and coherence of forms to generate a reading order. The reader, in this order, tends to show preference for grouped panels over non-grouped areas, smooth paths over broken paths, not jumping over units and not leaving spaces without reading (2013: 95). Regarding the first panel to be read, as in most visual media, the preferred entry area is situated on the left upper corner of the page, the most upper one, or, if not, the one situated more to the left. Most readers also leave the panel through its exterior border and 1 In manga, as it is known, the right to left reading direction still dominates. The adaptation to the western reading direction creates some problems (see de la Iglesia 2016) and nowadays, also due to cheaper editorial costs, most manga are published in their original reading direction and orientation, usually including reading instructions for readers not familiar with the convention.
del Rey Cabero: Beyond Linearity Art. 5, page 3 of 21 advance in the reading towards the right. The experiment showed that the Z-pattern of reading is not always the only option; panel groupings prove extremely relevant, in a similar way as the processing of images is affected by visual patterns, as pointed out previously by the Gestalt perception theory. However, other categories have been proposed apart from Peeters's, Groensteen's and Cohn's, namely the ones put forward by Chavanne (2010). In his Composition de la bande dessinée, besides regular and rethorical, he adds semiregular compositions and the concept of fragmentation (where tiers are the fundamental unit, not the page).

Linear and holistic reading
The relationship between panels (and images in general) and the page is essential to comics as a medium. Two apparently opposite activities join within the physical space of the page: seeing and reading. Numerous critics have pointed out this foundational tension of comics: linear against tabular (Fresnault-Deruelle 1976), story (récit) against scene (tableau) (Peeters 2003: 48), lineal against holistic or sequence against surface (Hatfield, 2009: 139), and macroreading (macrolectura) against microreading (microlectura) (Arredondo 2015: 196). Whatever name is given to these two elements, they refer to a contrast between time and space. The reader processes the images and can stop briefly or for a long time in one or a series of them, as when contemplating a painting. Moreover, in the case of a panel containing text, the reader will normally perceive the image first and then complement it with the reading of the text (unless the text is printed in a large font or highlighted in some way). Readers are aware that they must then continue and leave that panel behind in order to face another panel. The temporality of reading a panel is obviously variable and it does not only depend on the quantity of text or what is depicted on the image, but also on the location of the image itself on the page, the reading order or more subjective factors (each reader might stop more or less time on a particular panel). The same could be applied, mutatis mutandi, to the reading of the whole page.
However, it is quite difficult for readers to control the gaze and avoid seeing other adjacent panels to the panel they are looking at. This peripheral and complementary vision has been called peri-champ by Peeters (2003: 18) and del Rey Cabero: Beyond Linearity Art. 5, page 4 of 21 panoptisme by Samson (2012). If we compare this reading with the reading of a wordonly text, we soon realize that, even if we are aware of the words that follow or surround the one we are processing, we hardly ever read a whole sentence or a paragraph accidentally: a certain conscious effort is needed for this to occur unless the words are highlighted in some way (colour, size, etc.). This situation contrasts with the processing of comics, where the eye tends to wander by nature, finding meaningful units. As Cohn, Foulsham and Wybrow point out with regards to the results of one of their experiments, 'when presented with multiple panels, participants might not inspect the images in a strict order. They might skip ahead, flip regularly between earlier and later images, or build up an overall idea of the narrative before inspecting more closely' (2016: 568).
Images are thus normally located with more ease than words in a text (again, unless words are highlighted in some way). In this sense, comics seem to be less inclined by nature to maintain an exclusively linear reading due to peripheral vision and the possibilities of the holistic reading of the page. It seems, though, that writing in a succession of lines read left to right, top-to-bottom has turned out to be the preferential vehicle of the production of meaning and knowledge throughout history.
Nick Sousanis's Unflattening (2015) precisely uses the medium of comics as a way to explore the relationship between words and images, questioning the flatness and unidimensionality of verbal thinking by comparing it to the great potential of visual and holistic thinking. His work, which uses a different layout for every page, makes use of various visual metaphors to express this contrast. An example of such contrast can be found when text-based thinking is represented by dominoes advancing in a single direction and visual thinking is depicted by the simultaneous relationship of planets in space (2015; Figure 1).
The holistic potential of the page was very relevant from the very early stages of comics. In this sense, it is important to remember that comics such as The Yellow Kid (first published in 1895) used different formats throughout its lifespan. Such formats included the strip, but it could be said that the predominant form in the beginning was a single image in the page, without divisions, a sort of superpanel populated del Rey Cabero: Beyond Linearity Art. 5, page 5 of 21 by Mickey Dugan (the 'Yellow Kid') and many other characters. Although one of the most famous instalments is the one published on the 25 th October 1896, in which a hidden parrot talks through a phonograph, the series did not often include text through balloons, but rather through posters, signs and banners. Thus, the reader processed the image in a holistic fashion, focusing on the accumulation of various peripheral details and characters in a non-linear and anti-sequential reading.
Furthermore, even when using panels, some authors look for a pictorial effect through the so-called super panels, polyptychs or multi-panel pan sequence.   and, according to Linton (2015), arguably overused in the mid-1980s. Some of the most famous comics artists and graphic novelists (such as Alan Moore and Frank Miller) have also experimented with them, and today they are just another visual technique in the language of comics, which can be used discreetly to link a few panels or more noticeably on the page or the double page spread.

Multidirectional and multilinear reading
We have seen, therefore, how the holistic reading process in comics differs from the conventional reading of text. As Chute argues, One of the centrally interesting aspects of comics' narrative composition within the space parameters of the printed page is its ability to use that space to suggest different kinds of reading-models that aren't necessarily linear, or that do not actually embed a "right" way to read and to look, but rather offer a density-which is often a layering of temporalities that the reader then decodes (2012: 413).
The examples that we have provided above suggest that the density Chute refers to was present in the comics medium from a very early stage. Moreover, the linear reading was not so clearly established at this time. Some critics have pointed out the numbering of panels (see Witek 2009)  contiguous panels work as mirrors and the reader is forced to read the panels first from left to right, then from right to left, and finally horizontally but in two parallel rows of panels that occur simultaneously. This technique has been called ' overlap' (solapamiento) by Gasca and Gubern (1994: 638)

Translinear reading and the book form
The possibility of easily locating images in various directions is perhaps one of the great differences between reading a conventional text and reading comics. It is important to remark that peripheral vision plays a role not only within experimental or avant-garde work, but in all comics. Images, in spite of their individuality and separation from the rest, are ' conscious' of their common existence, as stated by the famous principle of iconic solidarity described by Groensteen (2007: 18-20). This coexistence is, in fact, one of the main characteristics of the medium: Images in print do not chase each other in the same way as scrolling on-screen images: they stay put, remain available, can be checked, compared and returned to multiple times. This is one of the greatest riches of comics: the almost instant accessibility of each of the moments that make them up (Groensteen 2008: 50). 3 This foundational characteristic of the medium is often used by authors, who employ various images created to be compared. These correspondences or 'visual rhymes' can be of a different nature (graphic, chromatic, stylistic, etc.), both based on repetition and contrast of the elements portrayed (panels, sequences, characters, objects, etc.).
This process of establishing correspondences, which is non-linear and goes beyond the simple relationship of cause and effect, is denominated 'braiding' (tressage in the original French) by Groensteen (2008: 50). These connections can appear both at the level of the page or the double page spread. Some authors might create some sort of cohesion on the spread (the same scene or temporality, for instance), while others do not give importance to this unit and focus instead on the single page (or even smaller units, such as the strip).
In any case, when dealing with the codex format, the reader always faces (except for the first page) the double page spread. As Spencer Millidge reminds us, The eye immediately takes in the whole page-or the whole spread-at once, so it should always be designed with that in mind. There is no way that the artist can physically prevent the reader from looking or glancing at the last panel before the first; but the reader has a tacit agreement with the creator to read the page in order. Even so, any big change of scene, new chapter, 3 'Les images imprimées ne se chassent pas les unes les autres comme le font celles qui défilent sur un écran: elles demeurent lá, disponibles, vérifiables, comparables, offertes à des aller-retours multiples.
Cette accesibilité quasi instantanée de tous les moments qui la constituent est l'une des grandes richesses de la bande dessinée.' Translation is mine. We see again the key role that peripheral vision plays in processing comics. There is no doubt that a tacit agreement between the author and the reader exists. However, as previously mentioned, the eye tends to wander, which can be very challenging for the reader. In strips or one-page stories, for instance, the reader might involuntarily see visual information from the last image, which could spoil a dramatic end or a final gag. That is probably why there is a certain convention in comics to build tension before the turn of the page (that is, in the recto page), a mechanism frequently used by authors such as Hergé in his Tintin series. For instance, in Le Lotus bleu (1946: 57-58), Tintin and his companions are about to be beheaded by Wang in the last tier of the recto of the page, but the tension is resolved with Tchang's intervention in extremis when the reader turns the page. The book as a vehicle has tended to be the container for a discourse where the story progresses linearly parallel to the succession of pages that inevitably move forward towards the last page. This conception of the codex format has been described as 'teleological narrative' by Masten et al. (1997: 2): 'the book, as a technological form, is organized to be read from page 1 to page 2, from page 2 to page 3, and so on to the end of the book. This is of course a possible way of reading a book, and one that was encouraged by the development of narrative fiction in the eighteenth century.' In this sense, the adaptation of the comics discourse to the book form is an essential step that is not often taken into account by critics. Indeed, the codex format provokes, as Baetens points out, ' an undeniable vectorization of discourse; the book consecrates a linear, or more exactly monovectorized, reading, that distinguishes (and sometimes discriminates) a start and an end' (Groensteen 2007: 147). Notwithstanding this undeniable influence of the codex in the linear discourse of comics, the codex can obviously generate other forms of reading.
Not only can comics readers examine images within a page before 'reading' it (strictly speaking), as we have seen, but they can also extend this mechanism to the whole work. This is another essential difference from purely textual reading. A del Rey Cabero: Beyond Linearity Art. 5, page 14 of 21 comic contained in a book is easy to glance through; when allowed, readers often do this to get an impression of what they will find (art style, page layout, etc.) before buying it.
As other critics have shown, readers always have a holistic vision at their disposal: they have the ability of seeing past, present and future simultaneously on the page. Bartual (2010: 15)   Captain Atom is both looking directly at the characters of the story and at the readers, who are handling the comic book itself. The character is also reflecting on the reading of the work itself: the time of discourse is altered and does not follow a chronological order, so the reader is forced to go back and forth to understand the events that take place. This is all reflected in the book format; the page, far from being a linear form, as it is traditionally considered, 'looks both forward and backward, toward the preceding page and the following, toward the front of the codex and the back, toward past discourse and future discourse' (Dagenais 2004: 62).
Therefore, the visual rhymes we mentioned before do not have to be limited to the level of the page or the double page. Indeed, some authors seek to establish con- One Soul (2011), by Ray Fawkes, uses an apparently simple nine-panel grid, but also deviates from linearity. This graphic novel employs the double page to create a macrospace of 18 identical panels recreating the lives of 18 characters in different historical periods. Each panel follows the story of a single character, turning black for the remaining pages of the book when he or she dies and being occupied only by text, a sort of monologue of the soul. Therefore, the reader of One Soul can complete one story through the pages or process them all at the same time (or any combination of these two options); in any case, it is probable that the reader will flip forwards or backwards through the book to connect these stories.

Beyond the page and the book
Finally, it should be remarked that the codex itself as a format is being questioned as a narrative vehicle for comics, with clear repercussions for the way in which it is read or engaged with. Various authors have created works that call attention to the comic as an object, linking it to other practices such as the artist's book (Drucker 2008  This article has explored how comics, despite inheriting a linear discourse from the codex, offer other types of readings. First, the page or the double page allows for a holistic reading that takes the space as a whole, even in more traditional grids. Second, the panels can be read in various directions (multidirectional reading) and can also generate more than one story line taking place simultaneously on the page (multilinear reading). It should not be forgotten, then, that both the page and the book as a form in the medium of comics do not necessarily foster a linear and conventional reading, as images in comics can be easily connected throughout the pages (translinear reading). Finally, some comics have proposed other formats (including