2.1Hypertext and hypermedia

The parallel presence of different modes of representation (text, image, animation, sound, video) tends to hide the fact that web content is mainly accessed by means of a single, unified organisational principle: by triggering references (i.e. links). The trigger, at present generally a click or a tap on a reactive screen, will increasingly involve other actions such as wiping and sliding owing to the rise of (multi)touch systems (cf. Machate, Schäffler, and Ackermann 2013: 19–20). The tool required is the web browser, which also gives data a visual representation by interpreting Hypertext Markup Language (HTML) tags. Other (often highly complex) processes performed in browsers are generally based on the embedding of (mini) programs (using, for instance, JavaScript and Ajax). As a result, users have a steadily broadening range of options, from selecting menu elements to choosing perspectives or engaging in “dialogues” with virtual “assistants” or “guides” (cf. Storrer 2004: 213), together known as “interactivity”.

A second means of access is the direct input of a uniform resource locator (URL), which provides the unique address of a website. It typically comprises the transmission protocol (“http[s]”), the characters “ː//”, the host name (e.g., “www”) and the domain (e.g., “example.com” ); these elements are followed by the rest of the file path (Gergen 2002: 367–369).

Especially in theoretical discussions, the term hypertext principle has been carried over from the early days of the web, when content was almost exclusively comprised of texts. In line with subsequent developments, it was augmented by that of hypermedia. In “hyperstructures”, content is distributed between linked modules; they are thus essentially non-linear, a characteristic that has important implications. The connections between these modules form a “network”, usually with a hierarchical structure, but which may also be more like a path or branch into individual sub-structures (cf. Storrer 2004: 221).

Seen from a linguistic perspective, hypertext is both fascinating and challenging. On the one hand, hypertext and hypermedia require a new understanding of text, which in its conventional form appears only on individual pages (and possibly in pop-ups). Even there, it must be tailored to the needs of web users (see Sections 3.5 and 3.6), whose complete, individual texts are constructed in the hypertext from the sum of the various elements they access, which may extend across several websites. This phenomenon, known as ergodicity, means that texts lose their unity and stability (cf. Marx and Weidacher 2014: 187, 191). On the other hand, in this text construction process certain criteria already identified and explored in text linguistics acquire a new importance. User behaviour is steered by a coherent “sense of purpose” which can evolve within pragmatic, hedonic or affective action frameworks (cf. Jakobs 2013: 132). Since the links contribute importantly to creating a text experience, they should be cohesively structured, or at least have the potential for cohesion.

Although the user’s path can to some extent be determined by the design of links, they still have ample scope to surf in various ways. Jakobs (2003: 243), for example, distinguishes between interest-specific, task-specific and preference-specific behaviours. In their analysis of methods adopted, Canter, Rivers, and Storrs (1985) classified access behaviours into: scanning (skimming through large areas of the web), browsing (following a path to a destination), searching (for a given target), exploring (the scope of a site) and wandering (random surfing).

In the early days of the web, the focus was largely on converting conventional texts into hypertext while striving to generate “added value” relative to the latter (Storrer 2004: 208). Now that the web has become part of everyday life, printed and electronic forms of publication and communication provide alternative options with different features for different purposes. In his comparative analysis of the websites of mid-sized enterprises in Spain and Germany, Sánchez Prieto (2011: 15–16) found that the German sites displayed greater distancing from print conventions than the Spanish ones.

2.2Communicative usability

First and foremost, the web is a technical product whose development has been and remains primarily driven by technicians (in the broadest sense). It therefore comes as no surprise that thinking about how to structure and design websites has become a distinct technical discipline known as user interface design (UID). The quality of interfaces in terms of their ease of use (in terms of ergonomics, effort required, accuracy and so on) is analysed and evaluated under the heading of usability. As regards language-related properties, qualitative requirements such as clarity, distinguishability, compactness, consistency and comprehensibility have been set (e.g., in various sections of the standard DIN EN ISO 9241). However, aside from some brief descriptions and a handful of examples, the ideas behind these keywords have yet to be operationalised (cf. Jakobs 2013: 128–130).

As sites incorporate more and more images, and more and more complex media mixtures, usability research is therefore increasingly reliant on openness to disciplines such as graphic design, cognitive science, semiotics, psychology, media sciences – and linguistics. Equally, since its “pragmatic turn” and the emergence of applied linguistics, this discipline has displayed a keen interest in technically-based communication. With regard to usability, it has contributed the concept of “communicative usability”, which emphasises communication modes, higher-level action frameworks, contextual conditions and the interplay of content, interface and support (cf. Jakobs 2013: 131–135). As a result, linguistics has been able to follow up on work in the media sciences, which were early to recognise that: “Web site design is communication design” (van der Geest 2001). From a number of case studies, this author derived a comprehensive checklist, ranging from strategic decisions to content, interaction and production processes. The items referring to text (van der Geest 2001: 148–149) relate to general features of communication, and so retain their relevance today.

2.3Architecture and links

Web content is mostly located “behind” the screen content actually seen by users, who also lack the tangible orientation provided by printed matter. Consequently, the way modules are divided up, arranged and linked is of paramount importance. Ideally, the user should be given an idea of a website’s structure from the moment they enter it, above all through skilful design of the homepage (see Section 3.2). Adopting a scheme devised by Blum and Bucher (1998), Storrer (2004: 218–219) defines three main approaches to modularising complex information. These are: by function (different modules achieve the same communicative goal by different means); by perspective (different modules provide different perspectives on the same object); and by topic (different modules address different sub-topics of a single topic). In structural terms, Kalbach (2008: 215–223) distinguishes between linear, net-like and (poly)hierarchical structures, as well as facets (where the position of an element is determined by the categories to which it belongs) and expanding structures.

The structuring principles applied are governed by technical developments (and by fashion). One current trend is an orientation towards the genuine and supposed needs of different user groups and roles. At a technical level, this is facilitated by “displaying” selected material from a general “stock”; in logical terms, various (partially overlapping) hypertexts are prepared in the background. In sophisticated systems that react “intelligently” to user actions, these are gradually compiled as necessary. From a user viewpoint, reception is simplified in that information irrelevant for certain roles is suppressed. Pre-modelling user identities so that every one of the legions of web site visitors can navigate successfully and also find the “right” information is not without risk and may call for target group surveys. The user perspective must also be anticipated; even choosing between such apparently distinct identities as business client or private customer can be hard in the postmodern world.

In general, paths through the hyper-levels lead through layers of increasing degree of detail (on the application of the same principle within text modules, see Section 3.5). Storrer (2004: 219–220), for example, refers to the variants as “thematic granularity” (main topic, sub-topic, side-topics), “part/whole granularity” (wholes are broken down into ever smaller parts) and “foreground/background granularity” (a topic is enriched by the addition of background information).

A site’s accessibility depends heavily on the configuration of the links it contains (to be precise, the anchors). These constitute an “Angebot [offering]” (Runkehl 2011: 148), or even a “promise” (Pernice 2014), and “have to convey both the motivations for users to activate them and the possible consequences” (Austaller et al. 2006: 101). Links should be self-explanatory, but in practice are often inadequate to make all potential consequences transparent. Austaller et al. (2006: 101) cite the following as some uncertainties that need to be overcome: “Navigation [. . .] Is the destination within the current web site or outside? [. . .] Download [. . .] Are tools (e.g., plug-ins) required to represent the document available? Process [. . .] Will it be possible to navigate back, or to undo the action?”

Any confusion runs contrary to the widely accepted notion that web users do not want their activities to involve much thinking (cf. Jakobs 2013: 120). Krug, who adapted this idea for the title of his best-selling book on web design (“Don’t make me think”), typifies user attitudes thus: “We don’t make optimal choices. We satisfice (satisfy+suffice, © Economist Herbert Simon)” and “We don’t figure out how things work. We muddle through.” (Krug 2014: 24–25). However, other authors suggest integrating a “desirable difficulty” in order to “use cognitive disfluency to slow the mind down” (Benartzi and Lehrer 2015: 212).

2.4Location and orientation

For the user to form and maintain a mental picture of a site’s content, they must feel certain of their current location within the hyperstructure (the opposite of feeling “lost in cyberspace”). One way of ensuring this is to organise effectively the space to be navigated while avoiding redundancy (one example of which is the “starshaped navigation” used in some indices, which requires the user to return to the index itself, even to call up the next or previous element). The larger the web site, the more important becomes additional help, a fundamental part of which is a constantly present display showing upcoming links and so giving the user a preview of the main surfing paths. To enable self-location and enable reverse navigation, the currently relevant section of the navigation menu should be highlighted. Additional path representations (“breadcrumbs”) are useful because clicking on their elements enables larger backward steps to key pages recently visited (cf. Austaller et al. 2006: 102–103).

Positioning within the overall structure is facilitated by a sitemap, which is an overview of the site’s hierarchical (or other) connections; again, clicking on the pages shown provides direct access. The value of sitemaps in web design is disputed. Critics regard sitemaps as unnecessary, arguing that properly thought-out links render them superfluous. Advocates view them as helpful, especially on extensive sites, because they can offer a greater depth of detail; the existence of a sitemap is also one factor in search engine optimisation (see Section 4.3, as well as Kalbach 2008: 65–66, 227–235).

By omitting the need to make choices, contents structured in pre-determined series allow the user providing cognitive recovery phases (cf. Handler 2006: 211). For certain sub-areas such as archives, chronological layering of the elements offered is the most appropriate principle of arrangement. Even in the case of modules structured according to other (e.g., thematic) principles, time stamps enable users to place elements in a timeframe (cf. Storrer 2004: 220–221).

2.5Web addresses

A site’s web address also serves a navigation purpose. Fixed by registration as a unique code, it constitutes the linguistic key to the site to be typed into the address bar of a browser. However, this is not the full story. For one thing, websites are also reached via individual links (bookmarks, advertising buttons and other web texts) and search engines. For another, web addresses are supplemented with extra functions on account of their prominent role. Web addresses have maximum effectiveness when they are easily memorised and thus ready to use at all times. They may make use of expressions already fixed in the consumer’s mind such as brand, product or other names, or slogans (e.g., coca-cola.com, imlovingit.com). Alternatively, communication strategies are employed to disseminate specially designed addresses (e.g., call.me, which reinterprets an existing domain ending, that of Montenegro), perhaps through large-scale advertising (poster) campaigns. When advertising and language come together, rhetorical (and other similar) effects that rely on structures that are eye-catching and/or easy to grasp (comcom.com, car4you.ch, tr.im) are sure to follow. Ideally, several objectives are achieved at once: attention is grabbed, memorability ensured and a message conveyed. For further discussion of web addresses in business, see Handler (2011).

3.1Websites as semiotic ensembles

The way in which web content is presented on computer screens today is essentially the result of trial and error since the emergence of the world-wide web and the emulation of best practice (or influential bodies). This process has been constantly dependent on software and hardware developments as well as organisational decisions about how sites are designed and “filled”. Websites are an amalgamation of building blocks. On the one hand, these are functional in nature. They serve to ensure orientation and progress (system-based, e.g., navigation pages and search interfaces), the carrying out of non-linguistic activity (task-based, e.g., order forms) and interaction with other persons (interaction-based, e.g., “contact”, visitors’ book). On the other hand, the thematic components contain the content, which may consist of company/product presentations, job ads, FAQ/frequently asked questions, etc. (cf. Jakobs 2003: 237).

Content management systems (CMS) are now standard practice. Increasing division of labour and specialisation is leading to a proliferation of roles such as CMS designer, responsive web designer, screen designer, content designer, e-publishing designer and web app developer to mention only a few. This has encouraged innovation in layout and design, and in some instances a conscious attempt to design sites individually, to break away from established norms. Yet there is also evidence of consolidation and maturity in the way sites are constructed. This is supported by the fact that users are generally loath to learn new logic every time they access a new web site. Krug (2014: 21–27, 64–70) demonstrates how strongly the perception of sites is affected by their form and appearance. At significant points, users always expect to meet the same elements (identification data, navigation, search interface, etc.).

This model is characterised by the following elements and properties that basically derive from traditional modes of perception (here, of cultures that follow a left-to-right convention) and the needs of navigation. First and foremost, constantly present elements should be distinguished from those representing new content. The former tend to include, at top left, a site identifier, normally the company logo. Krug advocates, in addition, a “tagline” directly connected to this; indeed he proposes (2014: 95–98) a succinct “tagline theory”. A navigation band running across the visible area gives a thematic breakdown of the entire site and allows direct navigation to the various elements (if necessary, via pop-ups giving more detail). In the top right corner the user typically finds tools allowing them to obtain necessary information (e.g., contact details, shopping lists) or to initiate actions (e.g., searching, ordering, changing language or, if necessary, signing in). Another prominent location for ever-present material is the footer. This usually becomes visible only after downscrolling and is thus a suitable place for content or processes which, while only occasionally used, are required to make the site complete (e.g., imprint, addresses, offer of a newsletter subscription). Partial or complete sitemaps (see Section 2.4) are also positioned here (cf. Kalbach 2008: 67), increasingly alongside links to social networks.

The left-hand screen edge is suited to local navigation, that is, navigation within the current level. The central area is thus left free for the display of actual content, with its strongly eye-catching upper portion generally containing images that appeal to the users’ emotions, as well as the site name. That leaves the screen’s right-hand edge as the most popular area for advertising (which has a life of its own and may contradict the above; see Section 4.1), for up-to-the-minute news or for job advertisements.

Given that overall aesthetic impression plays an important role in a site’s semiotic macrostructure, use is made in design of a grid (cf. Santa Maria 2014: 113–118). Of importance here are the distribution of visually powerful and lighter surfaces (or even white space), the balance between text and images, and the nature and combination of colour effects. Boucher (2013: 265, 271) provides examples of schemata for such zoning and for “wireframes” (schematic page renderings) on e-commerce sites. At the detailed level, the main objective is to present logically connected items so that they are also visually related. One way of achieving this is to call on gestalt psychology (cf. Boucher 2013: 46–53) with its principles of similarity, continuity, closure, symmetry, figure-ground and common fate (cf. Johnson 2010: 11‒23).

3.2The home page: A special case

The home page is not necessarily the page via which users enter a site: search engines and other links may take them straight to a particular (sub) page. Nonetheless, it has a special status because a start point is useful even, or especially in the context of a hyper-structure. It is therefore vital that this page can be accessed directly from every sub-page (via a “Home” button or, for instance, a constantly visible logo). Krug (2014: 86) offers an abstract model of a home page, divided up by areas and functions, which forms the basis of the following overview. Needless to say, the home page also features those elements that generally demand a permanent presence: identification details, global navigation, a search interface, registration (if applicable), footer elements and so on, and which the user must be able to locate there in the absence of such presence).

The home page stands out from the entirety of a site, and is accessed more often than its other pages; Krug terms it the “waterfront property of the Web” (2014: 87). It is hence also the location for very specific elements. Here it is possible to make explicit contact with the visitor (e.g., by means of a powerful welcome message) and to provide an overview of the site (with “teases” about its content, “[c]ontent promos” that “spotlight the newest, best, or most popular pieces of content”, and “[f]eature promos” which “invite [the user] to explore additional sections of the site or try out features”) (Krug 2014: 86).

Home pages have many other uses. Shortcuts placed there can provide the visitor with a direct path to especially important and/or high-demand content that would otherwise require several navigation steps to access. Regularly updating an area on the home page with new, topical content motivates potential visitors to return to the site more often than they otherwise would. Finally, its high visitor frequency makes this page a particularly lucrative location for advertising, especially that of a cooperative nature (cross-promotion, co-branding, etc.). More abstractly, the home page as a whole can be employed to foster credibility and trust by clarifying what the site does (and does not) contain and making it easy for visitors to identify rapidly the next steps they can take (even if these vary from one visitor to another). Furthermore, corporate marketing departments are becoming increasingly aware of the home page’s attention-grabbing value, as is apparent in the trend towards large alluring visuals. Sometimes these take the form of automatic slide shows to present the broadest possible range of images, now successively rather than alongside each other, and to encourage visitors to click on them. Elaborate navigation structures are invariably located in the lower area, with users clearly prepared to scroll there provided they have a clear overview.

Through targeted search engine optimisation (see Section 4.3), site visitors can be directed to a page which more exactly matches their search item than the home page. Similar considerations to the above apply to such “landing pages”, which are designed to trigger a desired conversion activity such as purchasing, downloading, form-filling, or “clicking through” to other pages (cf. Ash 2009: 31).

3.3Images versus words

As the carrier of visual information, the computer screen can tempt designers to make excessive use of graphic elements, in particular icons, and especially as buttons for (mostly functional) links. While icons have the advantage of utilising scarce space economically, they can also be difficult to decipher unless they are among the few genuinely established symbols, such as search, print and add to basket. This problem is exacerbated when the user is asked to select one of many icons positioned alongside or below each other as these will be perceived as tiny, barely distinguishable patterns (cf. Boucher 2013: 133).

For these reasons, it may well be advisable to use words rather than images in links. Words convey meaning directly provided they are part of the user’s vocabulary. Since websites are founded on information technology, there is a risk of employing specialist technical terms that mean nothing to many visitors. However, it is possible to build on those terms familiar to any computer user. Where there are valid reasons for an extensive use of icons (e.g., shortage of space), mouseover text boxes can clarify meaning and simplify usage.

One aspect of the user interface design debate among major software suppliers, for whom graphic design contributes to product distinctiveness, also impacts on the web. This is the promotion of a minimalist, “flat” interface for links in reaction to skeuomorphic representations (i.e. simulations of reality) sometimes perceived as awkward . In this way, however, links can easily lose their “affordance”, that is, their ability to make an object’s purpose visible (cf. Boucher 2013: 63–69; see also Section 2.2). Links should be expressed in concise, meaningful language; metalinguistic instructions (e.g., “For xyz click here”) are considered over-complicated (Redish 2012: 259–270). Lack of transparency can also be counteracted by contextualising links externally, by grouping them or adapting them to conventions (e.g., placing them at a specific location). Yet many companies deliberately reject excessively “flat” design, such as Google with its “material design”.

The proliferation of images begs the question of how they should interact with text. Pictures are deployed because they can communicate in ways that are not possible with language; at the same time, words remain a key element because images cannot do everything. Ideally the two will work together, with each type of code contributing its own properties as required. Stöckl (2011: 48–49) draws a detailed comparison between the two types of sign. Here are examples from each of the four areas he analyses. First, from a semiotic viewpoint, images are iconic and thus easily perceptible, while language is arbitrary and so harder to perceive. Second, images are perceived simultaneously and holistically, language gradually and in a linear manner. Third, the semantic potential of images is vague and indeterminate, while that of language is (or at least tends to be) precise and definite. Finally, in pragmatic terms, images portray well the relative locations of objects within a space; by contrast, language can explain the logical connections between elements.

In considering the relationship between specific configurations of text and images, and its effect on perception, it is assumed that the two should be melded to form a coherent communicative act. An important indication of how this process might be modelled was provided by Barthes, who was primarily concerned with the semiotics of images themselves. He described that which is depicted as a “message without a code” (1977: 36) and, on another level, identifies connotations deriving from the wealth of cultural experience. As regards the combination with text, he addresses three basic relations: text that specifies the message of the image (“anchorage”); the reverse situation, where an image supports understanding of the text (“illustration”); and a category where both contribute in roughly equal measure to conveying a message (“relay”).

More recently, text-image research has strived to refine its anayltical methodology. Bateman (2014) provides an overview of the state of the art. The list of chapter titles alone indicates the number and the diversity of approaches: “Multimodal relations modelled on account of cohesion”, “[. . .] modelled on grammar”, “[. . .] modelled on discourse semantics,” “[. . .] modelled on accounts of rhetoric” and “[. . .] based on speech acts, interaction and action” (Bateman 2014: 151‒163). These approaches also reflect the fact that relations between text and (moving) images are becoming ever more complex and must be embedded in entire text-image networks. In many cases, the relation of a single text to a single image represents just one aspect of such a configuration, or is found in a conventional hypertext module. Text-image analyses work best on the basis of concrete corpora as demonstrated by (amongst others) Stöckl (2011), who combines and adapts different approaches in order to analyse the case of advertising. Clearly, it is impossible to make general statements about text-image relations in websites; here, too, closely-defined corpora give the most satisfactory results (see Section 4.1: Runkehl 2011 on banners).

3.4Text visualisation

The aspect of written text that most immediately strikes the user and shapes his assumptions is its visual form. Adequate size and contrast with the background are essential for clear visibility. Surrounding visual noise can be confusing (cf. Johnson 2010: 41‒43). For longer on-screen texts – “[t]ype to live with” (Santa Maria 2014: 62) – simple fonts have become standard. Serifs (small dashes attached to letters) can be distracting, unlike in printed media, where they aid legibility. Italics generally appear uneven when represented in pixels and are thus to be used sparingly, as is also the case in print (and for emboldening). Underlining was originally used on the web to indicate hyperlinks but this conventio, being highly obtrusive, has now been largely superseded by coloured text (traditionally blue). Other secondary aids to recognition are: manageable line length, which reduces the length of lateral eye movements; left justification only because an unjustified right-hand margin makes for easier readability than the highly variable word spacing caused by full justification (cf. Crystal 2010: 203); and ample division into paragraphs, if possible with subheadings.

Such headings, as well as other short but large-font word sequences – “[t]ype for a moment” (Santa Maria 2014: 59) – may demand non-standard typographical choices. In such cases, semiotically “charged” fonts (decorative, playful, strong, minimal, etc.) may be used to convey the particular character of the content concerned. In the case of navigation categories, for the designation of which there is generally little space, neutrality and rapid perceptibility retain paramount importance. Limiting the number of fonts used (the recommended maximum is usually two) can prevent the screen becoming a typographical jumble. Another issue is the combinations used: depending on the style to be conveyed, the aim is to achieve an appropriate blend of differentiation (no fonts that are too similar) and harmony (strong complementarity). “[A] good rule of thumb is to pair a serif and a sans serif.” (Santa Maria 2014: 78). Colour, size and spacing can be employed to generate further variation within an otherwise consistent display.

As regards background coding, the current trend is towards logical renderings, such as <strong>, which leads the browser to apply its own highlighting mode, rather than <b>, which always results in emboldening. This approach need not imply losing control over the final on-screen appearance. The preferred solutions for such instructions (e.g., the way <strong> should actually appear) are cascading style sheets (CSS). These enable structural logic and final appearance to be processed separately, even though they are combined in the browser.

3.5Text composition

Most commentators agree that longer texts on websites are inadvisable (cf. Price and Price 2002: 85–95), with the exception of, for instance, online newspapers, blogs and scholarly sites. According to web specialists, text should be presented according to the principle of progressive disclosure (i.e. progress from the general to the particular). However, this requires qualification. The principle of the inverted pyramid borrowed from journalism is by no means universal (cf. Jakobs and Lehnen 2005: 180); it may apply to news, but not to instructions.

Textuality on websites typically displays a high degree of pre-determination owing to background (software) technology. Even the earliest HTML codes (e.g., <h1> to <h6>) aimed at strong visual segmentation through (sub)titles. Yet simply “[g]oing through existing content and putting in a heading every so often does not produce good information” (Redish 2012: 168). Continued expansion and, above all, CMS systems in which key decisions on text structure are taken in advance, at system management level, have led to further widen the range of ready-made text modules. One classic example is the FAQ section; initially a straightforward list of questions and answers, this now often displays the questions only, the corresponding answers being revealed by clicking. The same principle has been transferred to the text level in the form of an “accordion”, in which, for instance, subheadings are expanded into longer texts by clicking. Another variant on text segmentation is to place some elements (e.g., supplementary information) in separate boxes. For compact content providing clarification at a specific point (e.g., in running text or on a map), layers make efficient display formats, above all because the main page remains visible (Redish 2012: 140).

Here, linguistics raises the question of text types. Regarding these as communicative devices associated with particular functions highlights their close connection with technical structure. Jakobs and Lehnen (2005: 166) stress that hypertext types tend strongly to be polythematic, that is, to combine several sub-topics under one thematically overarching roof. The result is a nested structure, in which entities can be integrated at various levels. From a general perspective, smaller units tend to form building blocks, although they can also be classified as text types. On the basis of practical knowledge, scenario-based entities such as virtual shops (cf. Handler 1998: 144–146) can achieve strong pragmatic coherence (and also create expectations about the way the site operates). Given the very different communication configurations involved, it seems sensible to draw text-type boundaries between websites, newsletters, (micro)blogs, video portals, social platforms and so on, while establishing interconnections between the types. Conversely, hybrid forms of usage may be identified within a single type, e.g., the WordPress blog service, which is also used to create websites.

One technical special case is that of pdf documents. Not in principle part of the display level of websites, these can be regarded as de facto print texts that are distributed via the internet. Redish (2012: 120–121) sees the use of pdfs as justified in situations that demand the form of a printed document but where transmission via the web is chosen for cost reasons or because access is more convenient for users (permanent retrievability, no physical effort). However, pdfs are not suitable for displaying large bundles of information, of which only specific details may interest the user. Moreover, smartphone displays are totally unsuited to pdfs.

3.6Language properties

The ever-present danger that surfers may leave a site with a single click demands that the style be carefully chosen to offer a user-focused reading experience. Admittedly, reader-friendly writing was desirable long before the internet. It is part and parcel age-old considerations about the quality of language and communication which have crystallised in the form of prescriptive style guides, one example from the English-speaking world being “The Elements of Style” (Strunk and White 1975). Even in the new-media age its reputation as a classic has been maintained by allusions to its title, such as “The Elements of E-Mail Style” (Angell and Heslop 1994) and “The Elements of User Experience” (Garrett 2003:).

Unsurprisingly, general advice on good writing style is also applied to the web, as in these examples: “Write short, simple sentences”, “Cut unnecessary words”, “Put the action in the verb”, “Use your site visitor’s words”, “And always use plain language” and “Write in the active voice (most of the time)” (Redish 2012: 198, 224). At the same time, linguistics aims to provide research-based guidance while qualifying over-generalisations. The passive voice, for instance, is suitable for expressing negative or unwelcome messages. Compare, for example, “Supplies of part X have been discontinued” with “We are discontinuing supplies of this part” (cf. Jakobs and Lehnen 2005: 180). Where dialogicity is involved, interaction style remains relevant. Here Redish (2012: 202) recommends using the imperative, directly addressing the reader as “you” to ensure gender neutrality and referring to the company or organisation behind the site as “we”.

In view of the large number of text types, hardly any general characteristics can be identified. Attaining even such apparently obvious aims as “clarity” in practice is heavily dependent on users’ status. Whereas for a subject specialist a precise technical term will probably make the meaning clear, the layperson may be better served by paraphrasing. For interdependencies with corporate language, see Chapter 27.

3.7Alternatives to text

As part of the intensive consideration of presentation style so fundamental to website communication, it is necessary to ask whether some types of content routinely transmitted in sentence form could be better represented on the web in another format. Lists are an obvious choice for elements in the same ontological or pragmatic category. Augmented with bullet points, these can contribute to establishing the site’s visual identity; numbered, they can highlight the order of processes and instructions. Lists require careful formatting; the spacing between elements should be greater than between the lines of a single element, while parallel syntactic patterns can assist cognitive processing.

Given the potential of tables to establish relationships, they are especially well suited to if-then structures and to providing answers to series of questions (e.g., in FAQ sections). Tables are also the classical location for comparative figures. On the web, they can be made especially clear by the use of background colours to emphasise their structure (e.g., two alternating colours for the rows). Their width should be limited to the screen size to avoid the need for horizontal scrolling (cf. Redish 2012: 227‒251). Infographics are well-suited to illustrating pragmatic relationships and processes (see Chapter 4), while images are advisable where their strengths match communicative goals (see Section 3.3).

4.1Advertising

The phenomenon of advertising permeates the web at almost all levels. In general terms, it is apparent as commercial policy in the form of business models such as Google AdWords (paid advertisements on the margins of search result lists, in an almost identical design and with links). Innovative devices (e.g., fading in the address of a nearby restaurant on a geolocalised device via Google Maps) can achieve the purpose of advertising relatively easily; it suffices to display the name and address on the map (although logos, images, links to company websites, ratings and virtually any other information can also be displayed). Traditional distinctions between marketing forms are cast in doubt as they may be intermixed or just one click apart, and so truly become parts of a single hypermedia construct.

The form of advertising recognisable to website users as such – “display ads” generally consisting in graphics, mostly with a language element – may be static, animated or interactive (Janoschka 2004: 52–62). Although some such messages do not include links, most do. For links are the ideal means to exploit the hyperstructures typical of the internet and so to lead the viewer on (to a company site, to additional content on a specially designed site or directly to an order form). A distinction is drawn here between the “‘initial advertising message’ [and the] ‘linked advertising message’ [. . .] with additional links which lead to further internal or external target places [. . .], the ‘extended advertising message’” (Janoschka 2004: 50). Particular problems are posed by variants such as pop-ups and fade-ins, which are invasive and may force users into further, undesired actions. Also classifiable as invasive are graphics where a simple (and often accidental) mouseover while surfing triggers additional effects such as audio messages. Pop-ups can be blocked by special software or using browser settings. In their counter-lobbying, the web industry and operators of internet services financed by advertising argue, for example, that they are dependent on advertising revenue. Pop-up blockers that simultaneously maintain whitelists for paying companies constitute a particularly perfidious business practice.

The term “banner advertising” derives, by analogy with outdoor advertising, from the most popular form of advertising space. In the broadest sense, this kind of advertising includes variants whose names in specialist jargon reflect their shapes, such as “(standard) banner”, “tile”, “leaderboard” and “skyscraper” (Maddison Multimedia 2015). Currently, framing and layered ads are in vogue. In this chapter we do not concern ourselves with purely visual images; instead, we concentrate on the messages conveyed by language or language-image combinations.

A study of classic banner advertising by Runkehl (2011: 293–305) demonstrates that this is steadily moving closer to “traditional” configurations (with a headline, slogan and logo) as advertising space increases, partly owing to desktop’s ever-larger screens. More text is also being integrated into less space in the form of “animated advertising message[s]” (p. 174), that is, banner cycles. Temporary deictics (e.g., “now free of charge”) are another special feature; the desired rapid user response to them is also necessary because banners may change in random sequence. Moreover, text clearly predominates over images. As we are dealing with graphics, typography is not bound by the usual font requirements; re-shaping (even distorting) a font can elicit associative responses and, in interaction with text and images, create original messages (cf. Ortner 2015). In Runkehl’s corpus, these opportunities are taken up only to a limited extent. Janoschka (2004: 132–158) is especially interested in the linguistic means (e.g., questions, imperatives, personal and possessive pronouns) used to persuade users by integrating them communicatively and the role played by emotional modes of address (e.g., via trigger words or simplifications). However, involvement devices go beyond the purely linguistic and include devices such as fading in a “virtual smart agent” at strategic points in purchase decisions (cf. Veloso 2013: 182).

Given that electronic screen displays can be changed so easily, it has become popular among marketeers to integrate eye-catching corporate design components from advertisers (e.g., the colour magenta for T-Mobile) throughout a web site such as a daily newspaper. One aspect of the general trend towards personalised advertising is the introduction of procedural flexibility to websites by means of context-sensitive content advertising in the form of so-called native ads that “promote engaging content with stories users want to read without changing the form and function of a website” (Native ads 2015). The ever-expanding range of web tools offer consumers involvement, sometimes with weighty – and desired – consequences: web site visitors who accepted an offer to use an image-editing application in order to “age” their photos on the web site of a financial services provider were more likely to buy a pension-related product (cf. Benartzi and Lehrer 2015).

The extent to which differing interests compete on websites is apparent from the fact that page views, one of the metrics of online advertising, are sometimes favoured to the detriment of readability, with content divided into several modules to generate clicks. This has implications for text, which must incorporate, at critical points, strong incentives to click ahead. Nodder (2013) has focused on devious ways “to lead [web users] into temptation” with internet advertising.

4.2Web design for mobile devices

Since many websites are now accessed via mobile devices with relatively small screens, it is essential to tailor content to this type of reception. Initial approaches such as the creation of special “light” versions with limited content (and cross-references to the desktop version if more detailed information is required) have proven unsatisfactory. Unless adequate strategies are sought, numerous problems arise: confusingly truncated titles and texts, arbitrarily cropped photos, informative diagrams shrunk to the point of illegibility and tables reduced to a jumble.

The solution, so-called adaptive content, can also improve an existing desktop site. It is both organisationally and technically challenging, requiring as it does a powerful CMS system. Small chunks of content are arranged into a pool (e.g., title, subtitle, summary and running text, ideally segmented with subheadings), augmented by meta-information on status, and the whole assembled appropriately for the frontend device in question (cf. McGrane 2012: 47–82). Initially, content is generated in presentation-independent form; only later is it converted into text using devicespecific layout information. Formatting and selections are applied to specific screen situations, as well as to services such as blogs, Facebook, Twitter, search engine result lists and the like.

The ultimate goal of rendering all content suitable for mobile devices may demand the preparation of alternative versions to suit reception conditions, e.g., shorter titles with a meaning as close as possible to the original, or graphics displaying portrait format data shown in landscape format on a desktop.

Of course, smartphones and tablets also present different challenges as regards navigation. Since the home page remains extremely valuable as the first point of contact, it would be unwise to give it over entirely to navigation buttons. Where there are very clear user priorities, though, the most important buttons and fields can be positioned here (e.g., input masks for ticket bookings with Lufthansa); a second step is then needed only to access other pages. Increasingly standard use of an icon with three horizontal bars (also known as a “hamburger”) to denote navigation is helpful as it is widely used and understood. The time-consuming business of struggling through content on smaller screens (as more steps are generally needed) makes progressive disclosure a guiding principle in the packaging and distribution of content (cf. McGrane 2012: 111; see also Section 3.5). Given the responsiveness of modern smartphone displays, the elementary (and highly ergonomic) action of swiping has made even longer texts relatively easy to read – more so, in fact, than scrolling. Admittedly, it is impossible to obtain a proper overview of a longer text on a small screen, but the accordion principle can help to compensate for this.

4.3Processing for search engines

The number of visitors to a website is a highly important measure of business success. To a considerable degree, site access is initiated through search engines (Google in particular). It is therefore critical for a site to appear as high as possible in their results lists. Only in this way do sites have a realistic chance of being accessed; more than 90% of all clicks from Google are made on sites listed on the first page of results, with those near the top claiming more than 30% (cf. Chitika Insights 2013: 5).

Amongst other things, Google’s complex algorithms – the details of which remain secret – respond to links from other, ideally reputable sites (Larry Page’s “PageRank” principle that made Google great), the topicality, scope and frequency of content updates, the range of media employed and, more recently, responsive web design. Henze (2013) lists more than 200 factors that can influence ranking, many language-related – as stands to reason since search queries are couched in language. Taken together, the criteria applied by Google and their weightings embody how the search engine defines relevance for its visitors – although its own interests also likely play a role. At the same time, companies are invariably keen to figure high in hit lists, and therefore make use of search engine optimization (SEO) to make their websites responsive to known and presumed factors used in search algorithms.

Strategic areas where the emphasis is on language include, for example, everything written in the HTML code <h1> to <h6>. These are the tags that create the hierarchical structure of text, from the text heading and lead paragraph to the subtitles and subheadings (up to six levels in total). The linguistic material contained in these levels – and in <title> (page title) and <meta [. . .] description> – is one of the preferred targets of search robots. The more clearly it expresses a site’s content and its components, the more accurate the hits. The first 50 to 60 characters of <title> become the first line of entries of Google search results lists and recur in the top line of the browser frame, while the first 150 characters of <meta [. . .] description> provide the description that largely determines whether or not the user clicks on the result concerned.

Google is also geared towards <strong> tags, towards any language associated with images (from the file name itself to alternative text in <alt> and to captions) and towards highly expressive URLs. Effectiveness is thus enhanced by distributing keywords characteristic of the site throughout these locations. Running text must be reasonably long (at least 200 words) to ensure that search engines can comprehend content properly; responsiveness is further improved if it includes key terms (several times if possible, but not too often). From an SEO perspective, Andrieu (2014: 113–116) recommends creating a separate page for every relevant concept, with keywords placed at the locations mentioned above. In view of reports that Google algorithms increasingly incorporate semantic aspects, it also seems advisable to include a sprinkling of semantically-linked words (synonyms, hyponyms and hypernyms).

Since there are so many intricacies to be considered, SEO is becoming a specialist area and occupation in its own right. Attempts are also being made to circumvent web robots. In the internet’s “Stone Age”, when search engines simply counted keywords, large numbers of these would be “hidden” in sites as white text on a white background (invisible to users but identifiable by web crawlers). Such simple deceits are uncovered immediately these days; both the strategies used to fool search engines and the means of defeating them have become much more complex. Generally, it should be remembered that SEO should serve rather than undermine a site’s communicative purposes. Hence McGrane’s advice (2012: 118): “Write two different versions of your headline. One that’s designed for human readers, who appreciate style, humor, and even the occasional pun. Write a different headline that’s crammed full of SEO-friendly keywords, and let Google [. . .] chew it up and spit it back out.” Yet a Google algorithm may already exist to deal with this approach!

4.4Evaluation, culture(s), ethics

This section will consider – for space reasons only briefly – some important issues not yet touched on. First, testing and evaluating are essential to website implementation. Textbooks devote ample space to them (cf. Krug 2014: 110–141, 295–347), with Jakobs (2013) among others providing a linguistic perspective. Evaluation involves, once again, considering the extent to which access is ensured for users with particular needs.

As a world-wide medium, the web must also cope with the implications of cultural diversity. This can mean taking account of the culture-specific interpretation of colours, while cultural considerations must also be applied to layout, animation, sounds and image selection. Focusing on global presence requires a combination of global templates and localised provision, in which the communicative tone must correspond to the relevant languages or dialects (cf. Meidl 2014: 21–30).

Ethically speaking, it is necessary to ask “À qui profite le clic? [Who profits from the click?]”, as do Benabou and Rochfeld (2015), who trace values created in the attention economy and suggest fairer alternatives. Pasquale (2015) issues a wake-up call with the finding that we are living in a “black box society” in which we have less and less understanding of how the mechanisms that shape our lives actually work.