6.2.3　A working framework of the analysis of cohesive devices in the layout in the current study

As summarized in the previous section，the two aspects of layout analysis that the three frameworks all acknowledge are the structure of document and the typographic features for salience which indicate importance in meaning.They correspond to global coherence relations and local coherence relations respectively.That is，surface cohesion includes the devices in the layout that signal the coherence relations，including global and local，among the semiotic elements in the discourse.The working principle for layout analysis used in the current study will be developed around the two aspects.

Signaling devices for the structure of the multimodal discourses are an important aspect of cohesive devices.By way of structure，layout elements are progressively grouped into larger elements，building up a hierarchical structure with the entire page This view can also be accounted for from the perspective of psychology.Structure，in many cases，means regularity.In longer multimodal discourse such as the English learning leaflets and brochures，there are various patterns of regularity in the globally hierarchical structure.For example，the stages identified in the English learning leaflets form parallel relations under the overall discourse topic.According to Sanford and Moxey（1995），regularity induces coherence.They point out that parallel structures are a good example.Because parallel structures are regularities which can occur in varieties the bounds of which are as yet uncharted，they can provide a sound cohesive mechanism from which coherence emerges.Although they clarify their point by an example of a purely verbal text，they extend their conclusion to more general sense of regularity，saying：

Regularities of any kind may contribute to the processor's assessment of coherence.That is，the detection of pattern match，and more generally，regularity，may be the heuristic mechanism for accepting a“satisfactory level”of coherence.（Sanford and Moxey，1995：182）

The most important resources for the realization of the layout structure is“grouping”in Kostelnick and Roberts'（1998）terminology，or“framing”in the sense of Kress and van Leeuwen.As mentioned above，it is an essential part in the construction of coherence in a multimodal text.As Kostelnick and Roberts argue，grouping comes from the Gestalt principle which stresses the essential role the whole form plays in understanding its parts.They observe that，“by threading these parts together into manageable units，grouping enables readers to sort through the parts of a document more efficiently.Grouping creates visual cohesion;it's the glue that holds the parts together”（Kostelnick and Roberts，1998：61）.They identify three most important ways in which visual groups are created：likeness of form，spatial nearness and division.The term“division”has the attribute of the term“framing”in Kress and van Leeuwen（1996）.However，“framing”has a larger connotation than“division”.“Framing”covers resources for both connection and disconnection relations.The latter corresponds more or less to“division”.The analysis of layout structure by Bateman（2008）also mentions“framing”and“visual integrity”as two criteria of building up a hierarchical structure.The comparison of the three frameworks in this respect shows that，whereas similar resources concerning“grouping”are discussed under different names，Kostelnick and Roberts'（1998）proposal also includes an important device absent from the other two frameworks—likeness of form.Likeness of form，or visual similarity，is an important resource in signaling the parallel status of two layout segments，and thus a structural marker，and accordingly，a cohesive device.In summary，the cohesive principle in the structural line should include visual similarity，spatial nearness and framing.

Visual similarity can be achieved through the similarity of color，shape，composition，and others.Spatial nearness means that two elements are placed in spatial proximity to show connection in meaning.Framing covers both connection and disconnection.The former can be achieved by frame lines of various kinds and empty space，and the latter by vectors，linking lines，and other resources.

When discussing the graphic responsibilities of layout editors，Brooks and Sissors（2003：278）argue that the first in their list of responsibilities is that the layout should tell readers which stories are most important.The next several items along the list require the layout help readers know which pictures are related to which stories and which other stories are related to the one they are reading.The two requirements or responsibilities correspond to the two aspects of layout cohesion—structure and salience.Specifically，the requirements on the relation-marking can be regarded as one of the functions of the document structure，because in a complex document structure can mark the relations between major semiotic units and thus contributes to coherence.The first requirement concerning importance should be met through the distribution of visual salience.Since the previous texts in this section has discussed the specific devices for signaling discourse structure，the following will be concerned with the other aspect—visual salience.

Visual salience refers to the visual features of adjusting visual attention.The definition of salience provided in psychological and computational theories is very specific—salience refers to the physical，bottom-up distinctiveness of an object（Fecteau and Munoz，2006）.Different from structural markers which mainly work for a grouping effect among various elements in the layout，salience is a term that is attached to individual elements in relation to the surroundings.In other words，the salience of an element is a relative notion derived from comparison with its surroundings.It is a relative property that depends on the relationship of one object with respect to other objects in the scene（Fecteau and Munoz，2006）.There are many different physical qualities that can make an object more salient than other objects in the display，such as its color，orientation，size，shape，movement and unique onset.This view can find support in Kress and van Leeuwen（2006）who claim that salience“is not objectively measurable，but results from complex interaction，a complex trading-off relationship between a number of factors”（Kress and van Leeuwen，2006：202）.What factors should be grouped under this category has had no consensus yet.Kress and van Leeuwen（2006）list size，sharpness of focus，tonal contrast，colour contrast，placement in the visual field，and perspective.Kostelnick and Roberts（1998）emphasize the figure-ground contrast by various means，such as location on the visual field，color，underlining/framing，and other resources for highlighting.

It is important in the discussion of cohesive devices because salience is a criterion with which the importance of the various elements in a discourse unit can be marked.Inside each discourse unit，the elements are of different importance.Some are more relevant with the discourse topic than others.The salience resources visually mark the differences in importance，and through which help the readers find the important information quickly.

Because these proposals are made on the basis of multimodal documents in general，they need to be adapted to the special features in the English learning advertisements at issue.That is，the English learning advertisement as a genre have its own features in the expression of visual salience and accordingly，the importance attached.Moreover，since this book is devoted to the coherence between visual and verbal semiotic resources，those devices attached to only one semiotic mode such as the verbal texts（for example，type face）or the pictures（for example，sharpness of focus）will not be considered.Instead，we'll only focus on those devices that function on the interface of more than one semiotic mode，or on the composites which comprise both verbal and visual elements.With this constraint，we examine the devices in my corpus of the English learning advertisements that signal different degrees of salience and semantic importance.The finding is that，the devices that matter most and are most frequently encountered are color，size and location.The measurement of visual salience requires a comprehensive consideration of the three factors.

My proposal is supported by empirical evidence from eye-tracking studies.For example，in terms of spatial location，Wartenberg and Holmqvist（2005）have found that items at the top of the page are looked at significantly earlier than items lower down the page，and items on the left of a page are looked at significantly earlier than those on the right.Holsanova et al（2006）also verify that readers pay attention to items on the top of the page first，then the bottom.

The working framework for the analysis of surface cohesion in the layout of the English learning advertisements in this book is shown in Table 6.4：

Table 6.4　A working framework for the analysis of cohesive devices in the multimodal English learning advertisements

What needs to be noted is that，in actual cases of document design，these cohesive devices are often used simultaneously.They work in a collaborative way to realize the function of marking structural boundaries and creating cohesion.It is only for the convenience of analysis that they are listed as discrete items independent of each other.