version 1.0.1 – August, 2001
Leen Breure
Information and Computing
Sciences
University of Utrecht, The Netherlands
L.Breure@cs.uu.nl
Genre is widely used as classifying statement, especially in the field of arts, literature and media. A detective story, a novel, a diary or a newspaper article are each regarded as belonging to a different genre. Movies are labeled as SF, musical, thriller, Western, or horror (Lacey, 2000). However, this does not mean that generic classification is confined to printed material, the cinema and the TV. Since a few years the notion of genre has been applied to digital communication within organizations and on the Internet, not only for the sake of categorizing, but also to get a better understanding of the functionality, that electronic documents have within a community of users. From this point of view genre can be conceived as “an institutionalized template for social interaction” (Orlikowski & Yates, 1998), especially where verbal information is concerned. Genres constitute a framework for scholarly and organizational communication and are thus intimately linked to a discipline's methodology and professional practices.
As a consequence, the genre concept has a number of facets, which are the results from different theories in linguistic, literary, rhetorical and social disciplines, all engaged in studying this phenomenon. Describing genre from a sociocognitive perspective Berkenkotter and Huckin (1995:4) formulated a framework that summarizes the prevalent aspects of the modern genre concept:
The elements of this framework can be traced back in the theory of genre, as it has developed during the 20th century. The overview of this development below covers the dominant ideas and theories, that have given rise to the genre concept as summarized by Berkenkotter and Huckin. The first part is an outline of modern genre theory. It summarizes the historical background, necessary to understand the application of the genre concept to digital communication. The second part is a review of literature on digital genres (or cybergenres). This section is more detailed than the first part. The broader context is ‘genre as framework for electronic publishing’. This point of view is inspired by the idea, that genre creates shared expectations about the form and content of communication. In this way, genre characteristics are relevant to the design of electronic documents and websites, and genre analysis can be incorporated in the broad field of content engineering (or information engineering, as it is named elsewhere). Leading questions are, in which way such an approach might help to increase the effectiveness of electronic documents, and how the engineering process itself could benefit from a detailed analysis of generic elements.
The term genre comes from the Latin word genus and dates back to classical philosophy, where it was already used in the sphere of classification. Greek philosophy was concerned with questions about ultimate reality and the process of human cognition. At first sight this may seem a quite different subject, but remnants of Greek philosophy have entered into some parts of modern genre theory (Campbell & Jamieson,1978; Miller, 1984, see below), so it is worth to deal with it at some length. Plato's answer to the question of reality was his well known doctrine of Ideas or Forms, which asserted that the individuals and their qualities in the visible world were only appearances or reflections of a higher Idea, which was most real. For him, the objects of true science were in the world of Ideas.
Aristotle, however, abandoned many of the central doctrines of Plato, especially with regard to the question of reality. For him the individual perceptible thing was fully real. The entire visual world was constituted by substance and form. Form was the cognoscible element, that specified the individual and which could be abstracted from the objects in a process of perception. External objects impinged upon the senses, and due to the power of reason the mind was able to extract the essence (or form), which determined the nature of the observed thing. More precisely, in this reasoning it was not individuals in a material sense, but rather species, i.e. a group having some common characteristics, that had essences. A species is defined by giving its genus and its differences: the genus is the kind under which the species falls, and the differences tell what characterizes the species within that genus. For example, ‘human’ might be defined as animal (the genus), having the capacity to reason (the difference). For Aristotle, the universe was an strictly ordered hierarchy, so classification was inherent in this classical legacy [1].
The European Romantic movement of the 18th and 19th century (particularly in Germany) influenced the philosophic reflection on genre in different ways. It questioned a simple division of literary works into static genres as they had been defined in by the authoritive writers of Antiquity. It acknowledged the historical character of genre, altering across time, and it mixed this idea with Darwin's theory of evolution, as expressed in his Origin of Species (1859). The most pervasive legacy of Romanticism, however, was its doubt on the applicability of the genre concept at all. “Every poem is a genre unto itself” (eine Gattung für sich) is a famous statement by Friedrich Schlegel. Romantic art was conceived as self-expression, an ongoing process of pushing back frontiers, and this idea did not go together with any predefined scheme such as the rules of genre (Duff, 2000).
About 1900 this Romantic resistance to the harnessing power of genre classification is most prevalent in the work of the Italian philosopher Benedetto Croce. Every true work of art has violated some established kind and upset the ideas of critics. Thus they have been obliged to broaden the kinds, until finally even the broadened kind has proved too narrow, owing to the appearance of new works of art, naturally followed by new scandals, new upsettings and – new broadenings (Croce, 1902; 2000:27).
About 1920 reappraisal of genre started with the Russian Formalists, a school of literary criticism, led by Victor Shklovsky. They relied heavenly on the linguistic techniques of Ferdinand de Saussure and the Symbolist notion concerning the autonomy of texts. The Formalists sought to make their critical discourse more objective and scientific and emphasized the importance of form and technique over content. They considered literature as a complex unity of component parts, which could be analyzed as a system in a formal way.
Concerning genre, the Formalists elaborated the Romantic legacy in innovative ways. They kept the concept of evolution and viewed literary history and the development of genres as a dynamic process. A genre cannot be a static system because a new genre grows out of the consciousness that the old one has been supplanted by a new one. Literary evolution is discontinuous. At a period when a genre is disintegrating, it shifts from the center to the periphery, and something more fashionable takes its place. Yury Tynyanov (1924) regarded literature as a process of dynamic construction: some phenomenon becomes ‘automatised’ and evokes a counter-movement, which on its turn gives rise to opposing principles of construction. In addition to form Russian Formalists emphasized function as well. Both evolve across time. A new form appears, because an old one has exhausted its possibilities to fulfill its function (Duff, 2000).
About the middle of the century, the discussion was continued by a sociological movement, headed by Mikhail Bakhtin (1895-1975). After graduating from the University of St. Petersburg Bakhtin moved to Belarus, where he began to write his theories, which were critical to the Formalists. Because of Stalinist censorship, from which he did not manage to escape and which dispelled him to Kazakhstan and Mordovia, he published works under the name of others, like P.N. Medvedev. He is especially known for his publications on Dostoyevski.
Bakhtin broadened the field to extra-literary genres. Focusing on a single genre, the novel, he formulated an influential theory about speech genres. All human activity involves the use of language. Language is realized in the form of concrete utterances, oral and written. Although there is a wide range of individual peculiarities, some stable types of utterances are formed, which Bakhtin called speech genres. He made a distinction between primary and secondary genres. The primary (or simple) speech genres are directly related to verbal communication in everyday life. They enter the literary field, for example, in the form of dialogues in a novel. Secondary speech genres, such as dramas, novels, literary commentaries, arise in more complex and comparatively highly developed cultural communication and are primarily written. During their formation they absorb various primary genres. In each epoch certain speech genres set the tone for the development of literary language. For Bakhtin genres are not simply sets of rules and conventions, but ways of conceptualizing reality, forms of seeing and interpreting particular aspects of the world. They are connected with expectations about length and compositional structure. They are the fundamental molds in which we cast communication ( Bakhtin, 1979).
Modern genre theory is greatly influenced by North American genre studies and by a parallel movement in Australia, referred to as the Sydney School, after its main base in the Department of Linguistics in the Sydney University. Both recognize the importance of social factors, the primacy of the social in understanding genres and the role of context, but differ in other respects. For example, the Sydney School has been particularly interested in textual features using schemes of linguistic analysis, and has emphasized the prescriptive nature of genres and their static aspects. In contrast, Americans have underlined the dynamism of genres with central concepts such as interplay and interaction and has focused on the complex relations between text and context [2].
The Sydney School is characterized by systemic-functional linguistics and semiotics, and owes much to the so-called register-theory of the linguist Michael Halliday. Halliday defined a register as “the configuration of semantic resources that the member of the culture associates with a situation type. It is the meaning potential that is accessible in a given social context”. Register was used to define individual characteristics of a text as determined by its context. Related research has been directed to pedagogy and education. Jim Martin investigated genre in relation to student writing. He used the term genre as synonymous with register and focused on the semiotic function of texts within a social system . Other representatives, like Kress and Threadgold, formulated slightly different genre models, all positioning genre as expressed by linguistic text features in a wider framework of contextual situation and culture (Knapp, 1997:115- 128).
In the 80's Anne Freadman started to discuss genre outside the prevailing paradigms of linguistics and semiotics. Although her theory shows certain similarities with Halliday's notion of register, she postulates, that texts are the products of the interaction of a variety of ‘languages’, or semiotic systems, while in the Hallidayan model everything is reduced to the single semiotic system of natural language (Knapp, 1997:130). Genres can be, quite fruitfully, considered as extensions of speech acts. She draws a parallel between ‘discourse’ and ‘game’ and uses the analogy of playing and returning shots in tennis to describe the dialogic nature of genres. Genres tend to occur in pairs (the ‘dialogue’): brief and report, play and audience response, essay question and essay feedback. However, game can be a quite misleading metaphor as well. It might suggest, that text is the output of a set of rules (a ‘recipe-theory’), while practice is much more complicated. ‘Game’ and ‘playing the game’ are not the same. Playing the game is a ‘ceremony’, which involves a great deal more than the game itself, like background knowledge on the place and circumstances of the rules' apllication. Knowing a genre is also knowing how to take it up. What do you do with a form, if you've never been taught to fill one out? Genre is not absolute, it is pragmatic and turns as a game in a variety of social settings (filing, library classification, publishing and bookselling etc.) (Freadman, 1994).
A great deal of the American interest in genres is derived from the ‘New Rhetoric’ (the work of Burke and Searle, and Austin on speech acts), which shifts attention from structure and content of an essay to the contextual framework of society. However, the American philosophy is not homogeneous. Ideas from the New Rhetoric have inspired Carolyn Miller in her influential article ‘Genre as Social Action’ (Miller, 1984).
Miller's approach is centered around repetitive rhetorical situations, but she rejects the notion of genre as a recurrent pattern of forms used for simple classifications, which easily leads to reductionism and formalism. A classification should contribute to an understanding how discourse works, i.e. reflect the experience of the people who create and interpret it. Genre as action must take into account the context of the situation and the motives, the intention and the effect. In a discourse community a situation functions as a social construct with a well defined meaning. Genres are responses to it, typified rhetorical actions. She follows Campbell and Jamieson (1978), who postulate a Aristotelian fusion of form and substance (i.e. the semantic value) in rhetoric: “a genre becomes a complex of formal and substantive features that create a particular effect in a given situation” (Miller, 1984:25). Both are related: for example, the form of a text guides the reader and shapes his expectations and responses concerning content. In this way, form has a hierarchical relationship to substance. Leaning towards a few existing communication models, Miller proposes a ‘hierarchy of meaning’, each level providing context for a lower one. Genre appears at one of the higher levels. Form of life provides context to genre, while it is constituted by lower elements as episode and strategy:
| human nature |
| culture |
| form of life |
| genre |
| episode or strategy |
| speech act |
| locution |
| language |
| experience |
In a more recent publication she has further elaborated this theme, now regarding genre more in a reciprocal relationship with social context, rather than as a cultural artifact, to be positioned somewhere in the middle between the macro-level of culture and the micro-level of language (Miller, 1994:68). Following the structuration theory of Anthony Giddens, which deals with structuring of social relations across time and space, she emphasizes genre as that aspect of situated communication that is capable of reproduction, which is, of course, connected with its recurrent nature. The study of genres has to be related, therefore, directly to the analysis of the social structures (discourse communities), in which they function. In this view Miller comes close to Bazerman and Swales [3].
Bazerman has become well known for his study of the development of single types of texts through repeated use in similar situations. In Shaping written knowledge (Bazerman, 1988) he described the evolution of the scientific article from 1665 to 1800, from uncontested reports of observations and events, to arguments over results, to accounts of claims and experimental proofs. In a later study on patents, he charted the complex web of interrelated documents issued in the application procedures. In this context he introduced the term systems of genre to denote the interrelated communications. Bazerman's view of genre is rather instrumental:
“From the viewpoint of the participant in society, which we all are, I want to identify how the genres in which we participate are the levers which we must recognize, use, and construct close to type (but with focused variation) in order to create consequential social action. This machine, however, does not drive us and turn us into cogs. The machine itself only stays working in-so-far as we participate in it and make our lives through its genres precisely because the genres allow us to create highly consequential meanings in highly articulated and developed systems.” (Bazerman, 1994:79)
The same view of genres as institutionalized mediators between individual and society is supported by the applied linguist John Swales. He connects genre to a discourse community, which is characterized by a broadly agreed set of common goals, patterns of intercommunication among its members, and other social mechanisms that regulate membership. To further its aims, a discourse community maintains discoursal expectations, which are created by the genres that articulate the operations of the community (Swales, 1990:26). These purposes constitute the rationale for the genre. The rationale shapes the schematic structure of the discourse and influences and constrains choice of content and style. Swales' concern with genre is aimed at offering an approach to the teaching of academic and research English. His work demonstrates the general value of genre analysis as a means for studying spoken and written discourse for applied ends. In this context he relates genre to task, conceived as a set of goal directed activities relatable to the acquisition of genre skills (see glossary for his precise definition of task). Further elaborating this theme into a pedagogical direction, he formulates the expectations and actions characteristic for a genre in terms of cognitive schemata with associated procedures, referring to similar notions like scripts, scenarios, and routines, used in other theories. These shape our expectations and let us behave appropriately in a certain situation.We are prepared for a limited set of genres, already before we open a newspaper or a scholarly journal. So, genre as typified rhetorical action also applies to reading and writing.
Before continuing with the application of the genre concept to digital communication, we may pause for moment, and ask ourselves what elements of genre theory were available in the last decade of the 20th century, when the electronic publishing and multimedia documents became increasingly popular. The current online version of the Encyclopaedia Britannica still suffices with a superficial definition of genre: “a distinctive type or category of literary composition, such as the epic, tragedy, comedy, novel, and short story”. From the foregoing summary of genre theory, it may be clear that this reflects only the current, everyday meaning of the term. The genre concept has moved far beyond the simple notion of a merely static, classifying concept, based on a characteristic pattern in form and content. Documents cannot be easily classified on basis of a set of formal criteria without any awareness of their context and purpose. Dominant elements of various genre theories form the following short list, which is hardly more than a paraphrase of Berkenkotter and Huckin's description of genre above:
The growing digitization of communication and the dominance of Internet in the 90's have extended the field of study to digital genres. The Digital Documents Track of the annual Hawaii International Conference on System Sciences (HICSS) has become a frequently used platform for presenting results [4]. The following global themes emerge from the recent literature:
Studies in this category take the perspective that generic regularities in form and content arise out of a recurring communicative situation determined by the institutional context. They apply concepts of genre theory to discourse communities formed by the employees of organizations. In this way, genres show insight into the nature of the community in terms of work practices and interaction norms. Their establishment and evolution, measured, for example, by composition and use of the genre repertoire, are indicators for the dynamic processes in the group and organizational changes. Genre analysis may be useful to map workflow, or to model corporate information assets.
At MIT Wanda Orlikowski and JoAnne Yates applied the notion of genres to organizational communication. They studied, amongst others, the evolution from the formal business letter (through informal business letter and memorandum) to electronic mail, and analyzed the genre repertoire in the e-mail of a team of computer language designers, paying attention to the function of memos, language proposals, dialogs and ballots. In accordance with modern genre theory, they conceived genres as socially recognized types of communicative actions, and identified them by their purpose and by their common characteristics of form. The purpose of a genre is not based on an individual's private motive, but rooted in social agreement, also in an organization. “A genre binds the shared purpose to characteristic aspects of substance and form. Substance includes the topics, themes, and arguments, along with typical discourse structures, used to express the communicative purpose.” Form refers to three observable aspects of communication: medium (such as paper or e-mail), structure (text formatting) and linguistic features. Within the organizational framework, genres are part of a dynamic process of production, reproduction and modification. Individuals use genre rules to produce or challenge a genre (Orlikowski & Yates, 1994).
Genres tend to be linked or networked together in a way that constitutes a coordinated communicative process (e.g. a conference may start with a call for proposals, followed by abstracts and concluded with papers). Such a cluster forms a genre system and is useful for studying the interaction between people in a community (Orlikowski & Yates, 1998). By answering the questions Why, What, Who/Whom, When, Where, and How with regard to specific genres a taxonomy can be constructed, which may consulted for organizing work, as demonstrated by Yoshioka and Herman (1999, 2000): the right genre at the right moment furthers communication and decision processes. In this way genre analysis is an aspect of work flow design and planning. The appropriate use of genre ensures the information is received with “trust”, i.e. in accordance with the social rules and expectations. Digital communication require genres with rules and expectations that fit in with the general corporate work climate.
An other field of application is computer mediated communication (CMC) and computer supported collaborative work (CSCW) (Bergquist & Ljunberg, 1998; Erickson, 2000). Genre concepts interact with interpretive schemes about technology. When people begin to use new collaborative technologies, they are influenced by their knowledge of genres they are currently using and by cultural expectations, national or organizational (Yoshioka, Yates & Orlikowski, 2000). Lack of agreement on this point may cause hitches. Genre analysis helps to understand the communicative situation. Recently, genre has been defended as framework for gathering the requirements of a corporate electronic documents management system (Tyrväinen & Päivärinta, 1999). An approach based on genres would do more justice to the written communication actually taking place, not blocking emergent new types of documents as non-standard and not-supported (Karjalainen et al., 2000).
Carolyn Watters and Michael Shepherd emphasized the role of genre in the evolution of user interface for the Internet (Watters & Shepherd, 1997a), thus shifting the attention to the technical functionality of electronic documents. While genres in non-digital media are determined by the tuple <content, form>, cybergenres are characterized by the triple <content, form, functionality>. Normally, a user interface is designed with a specific users group in mind and with regard to specific goals and task. However, the Web usage is more fuzzy. The actual goals and tasks may vary considerably. Therefore, functionality can only be deduced from the purpose the author had in mind. (Shepherd & Watters, 1999). Because a genre carries with it a set of prescriptions and restrictions, it provides also an established combination of content and functionality:
“Digital genre provide a framework of familiarity for users of disparate systems. For example, if digital newspapers work like paper newspapers, readers understand the type of content expected and how to manipulate the presentation, with little if any training. When electronic dictionaries act like paper dictionaries and digital libraries like book libraries users do not need to learn what to expect of content, form or functionality. Users come to the system with basic skills from past experience and self-directed exploration can drive the acquisition of more advanced skills, as needed.” (Watters & Shepherd, 1997a).
They discovered, for example, that people preferred to read news in the form of a digital broadsheet, rather than as hierarchal linked one-story-pages (Watters & Shepherd, 1997b). This idea of genre as interface metaphor has been further elaborated by Elaine Toms and Grant Cambell (Toms & Campbell, 1999). A document provides various visual cues that enable users to grasp its form, purpose and function quickly. Correct identification occurs, when these cues are consistent with those defined by its genre. One can imagine, that, recognizing the genre, the user "loads" a framework of expectations, which helps him to interpret the document and facilitates a particular kind of use. The effect of genre on user expectation is also confirmed by Dillon and Gushrowski, who examined more than 100 personal homepages and asked subjects to select those elements that they thought should be included in any good personal home page. They found a broad agreement as to what a home page should contain (Dillon & Gushrowski, 2000).
Crowston and Williams (1997) were among the first to realize the importance of the genre concept for analyzing communication on the Internet. They noted, that the Web was an excellent place to study the development of genres, because of easy access and its inherent capabilities of experimentation, freedom of structuring, and interactions between many communities. They documented the range of genres in use on the Web by sampling and classifying randomly selected 1000 Web pages. Distinction was based on purpose, rather than on physical form. They identified 48 different genres:
| Archive item | Filmography | Problem set |
| Article | Geneology | Product information |
| Book | Government program description | Product reviews |
| Box score | Guide | Publication list |
| Chronicle | Home page | Ratings |
| Column | Hot list | Regulation or rule |
| Computer documentation | Index | Report |
| Concert review | Library acquisition list | Script |
| Demographic data | List of research projects | Server statistics |
| Directory | Meeting minutes | Source code |
| Discography | Memorial | Submission instruction |
| E-mail directory listing | Newsletter | Table of contents |
| Essay | News wire article | Testimonial |
| Faculty information | Order form | Univ. course listing |
| FAQ | Pamphlet | Users' manual |
| File directoy listing | Political party platform | Vitae |
| Type of genre | Count | % |
|---|---|---|
| Familiar genres | 507 | 60.6 |
| New, but accepted genres | 239 | 28.6 |
| New genres | 44 | 5.3 |
| Unclassifiable | 47 | 5.6 |
Most of the pages fell in the category of reproduced genres (about 60%, like FAQ, meeting minutes and course descriptions). Adaptation was mostly due to outgoing links, thus creating innovative variants of the traditional genre, e.g. geneologies using linking to display and navigate the usually unwieldy data in a family tree. Linking can enable a single page to serve multiple purposes and thus be an example of multiple genres. Among novel genres they rated homepages, hotlists and topical home pages [5], pages about web servers (like ‘Web site has moved’ or file directories), forms and pages that provided access to application (such as a trip planner). A small number of pages turned out to be unclassifiable. One could agree that they belonged to certain genres, but was unable to name them. Some of them might be considered as examples of genres in the process of adaptation to the Web: variants of an accepted genre, but with some features missing.
Following Yates and Orlikowski ( 1992), who suggested that novel genres are most likely derived form existing ones, they tried to explain the origin of the new kind of Web pages found in the sample. Although the antecedents of the home page is unclear, some organizational home pages bear some resemblance to company profiles in printed material. Hotlists seems unlikely to have developed from bibliographies (because of their dense format), but FTP sites and bulletin boards used to have similar lists of references. They concluded, that the pattern of linking contributed highly to the nature of the genre: a hotlist had only outgoing links, while an electronic book had typically internal links. Genre has implications for web site design. Designers may want to draw on accepted genre where appropriate and should be aware of the user's expectation of a genre. Although they should feel free to reject old forms that doesn't fit the new technology, they must be aware that new genres are often misunderstood or resisted (Crowston & Williams, 1999).
Shepherd and Watters and (1998) coined the term cybergenre to denote digital genres and devided them into two classes of subgenres: extant and novel (see diagram below).
They made a survey (Shepherd & Watters, 1999), similar to that of Crowston and Williams (1997), but at a higher level of abstraction and on a sample of much smaller size (96 web sites were randomly selected). Their purpose was studying content, form and function of Web documents. They identified (at this more abstract level) only six different genres:
Concerning the attributes content, form and functionality, these genres were characterized as follows:
| Cybergenre | Content | Form | Functionality |
|---|---|---|---|
| Home page | information about person/company | introduction hierarchical images animated images |
browsing |
| Brochure | products and services | shallow hierarchy high-impact visual |
browsing |
| Resource | subject-specific information | hierarchical images video audio |
browsing search discussion interaction |
| Catalogue | products and services | hierarchical images |
browsing e-mail ordering & inquiry search on-line ordering on-line enquire |
| Search engine | categories of sites URLs |
query box list of sites virtual document |
browse search |
| Game | challenge to user scenarios rules |
animation audio video scenes |
high level of interactivity collaborative computing |
Because the design of this survey differed in several respects from that by Crowston and Williams, they also mapped the results onto their genre list to make results more comparable. With some reservation they concluded that the Web might have changed considerably in two years time and that:
There has been some concern that the rapid advance in new technologies would have a detrimental and destabilizing effect. However, new forms require social acceptance. Transformation is always shaped by the interplay between technological and social forces. Because digital genres are rooted in the social practices of a discourse community, they will provide a new degree of fixity, comparable with the old stability procured by typographic processes. A genre reproduced with a new medium may be accepted or more or less modified. This transformation can be a long term process, in which technological innovations play an important role.
Simeon Yates and Tamara Sumner (Yates & Sumner, 1997) have demonstrated this for documents produced in software design. The novel capabilities of design tools made changes in genres more likely to occur. Because such technical documents had to be understood by stakeholders as well, these evolutions could not be exclusively technology driven. They required the a broader consent of the discourse community as well. Shepherd and Watters share the conclusion by Yates and Sumner (1997), that, in spite of the apparent evolution driven by technogical factors and changing functional requirements, the notion of genre still provides a certain degree of stability and fixity. The combination of content and form “provides a familiar and strong metaphoric reference for users that transcends changes in functionality” (Shepherd & Watters, 1998).
Notwithstanding this stabilizing genre awareness, some sectors may evolve relatively fast, such as online news. Eriksen and Ihlström examined the trajectories of evolution for three Scandanavian news sites from 1996 to 1999 and found considerable differences between initial plans and final outcomes. Starting with the idea of Web news as replica of the printed news paper, the news sites evolved into the direction of a truely electronic news services. By 1999 the Web media were characterized by fluidity and were largely used as primary outlets for hard news. Many of the characteristics of print media had been lost: the concept of an edition or issue had been given up, news sites were frequently updated, and news components “float” through the site. They concluded:
“From being replicas of the printed newspaper, the news sites are currently establishing a presence, design language and content of their own. The news stream of hard news and the archive of soft news being two design elements that have merged as specific to the web news genre. The data gathered for this research suggest that the current perceived role of news sites on the web is as hard news sources, stressing facts in favor of reflection. Stressing immediacy in favor of history.” (Eriksen & Ihlström, 2000)
As said at the beginning, the broader context of this overview is genre as framework for electronic publishing. Genre creates shared expectations about the form and content of communication and, therefore, is supposed to be relevant for the design of electronic documents. A document, here, should not be conceived as the digitized version of a text on paper, but rather as a container, which displays all kinds of content components, varying from text and images to animation, audio and video. A document-as-a-container has a user interface, which may be partly determined by the hosting software (such as a web browser). However, depending on the technology used, this interface can be built-in in the document itself as well, driven by internal programming code in the form of JavaScript, Java applets, Flash-scripting, XSLT-template rules etc. The document's generic form and functionality are related to this technical architecture and this is precisely the link between genre study and content engineering. Against this background two major questions are to be to dealt with:
Both themes can be broken up into a set of topics and can best be answered by commenting on the genre characteristics listed at the end of part I.
In the context of electronic publishing the characterization of a cybergenre as a pattern of communication sounds almost as a truism. Designing documents requires at least some implicit awareness of the audience and, therefore, will lead to some form of user modeling, which will imply a patterning of communication. Is the genre concept of any real help for the practice of information engineering? Does it provide a new point of view with added practical value? We do have all kinds of methods and techniques for developing websites and electronic publications, mainly stemming from usability engineering, human computer interaction and general system development methodologies – aren't they sufficient? How does genre analysis fit in with these conventional methods and techniques?
Most of these methods start from the paradigm ‘Know the User’. They rely for a great deal on surveys, interviews and group wise participatory design techniques and, in a smaller degree, on the analysis of the content used. The latter with a view to data modeling and data-based task analysis [7]. Jakob Nielsen starts the discussion on the usability engineering lifecycle with the advise to visit at least a customer site in order to get a feel for how the product will be used, to know the class of people who will be using the system (to draw up a users' profile), and to make a task analysis (the users' overall goals and their approach to the task) (Nielsen, 1993:73). Several more or less detailed variants exist, such as the following triad (Fucella & Pizzolato, 1998) [8] :
In spite of all the methodology, the practice of digital information engineering seems to be far from easy. For example, Melinda McAdams, who was involved in creating the online version of the Washington Post, wrote:
“I have spent the last sixteen months building an online version of the Washington Post. In many ways the experience has been like trekking in a wilderness without a map or even a compass, and all of us who have worked on the project have learned a great deal along the way. We have faced most of the hard questions about translating the newspaper to a new medium, and if we have not arrived at definitive answers, we have at least explored and grappled with multiple possibilities” (McAdams, 1995).
The major part of the problems encountered were not so much related to the user himself, but had to do with translating the newspaper to a digital medium. Some sections were hard to transfer to the new setting. Although the newspaper as a metaphor proved to be useful, especially for browsing, it was not of much help in actually organizing the digital information space. In a printed newspaper a reader can scan information. Online news must provide access to many times more information, while the screen display limits the visual space. This requires better search facilities and a different organization of sections.
It must be admitted that an online newspaper is not an average business website. However, this example highlights the core problem of organizing a quite complex set of requirements into a comprehensible and manageable information pattern. Here genre analysis could be effective, because it departs from a slightly different paradigm ‘Know the User's Life’: not just a user's profile with isolated tasks, but the interrelated whole of challenges and responses, that constitutes the relevant communication pattern. It shifts the focus from user characteristics to common practices in which the user participates, and thus provides a frame of reference for the third step in the user modeling process, i.e. information organization. Genre analysis does not replace user modeling, but is complementary to it. However, note that mere replication of an extant genre may not work (as McAdams' retrospective analysis makes clear) and that a thorough understanding of the web variants is necessary.
How to deploy genre analysis in an information engineering trajectory? Building on the work of Bazerman and others Paré and Smart have elaborated an analysis model, which appears to be easily transposed to a digital environment. They define a genre as a distinctive profile of regularities across four dimensions:
The first dimension, regularities in texts, refers to form aspects, like document structure, style, and formatting. The second, the composing process, covers a wide range of activities, starting with the initiating event, and includes information gathering and analysis (e.g. defining problems and forming opinions), individual writing and the technique of text production (like word- processing and publishing). The reading practices, the third category of regularities, refer to the way a reader approaches a text, how he negotiates his way through the text, how he constructs knowledge from it and how he uses this. Finally, the social roles determine what can and cannot be done by particular individuals and regards responsibilities, division of labor, and rights of access to information; due to these rules the genre is enacted in much the same way from one instance to another.
The main practical difference with the user modeling approach is the set of documents as starting point and center of orientation. Both methods characterize users and consider scenarios and tasks. The analysis suggested by Paré and Smart follows the information life cycle of creation, design, and use, departing from observable document features. This may be convenient in those cases were the audience is heterogeneous, it is difficult to interview potential users, or requirements from user modeling are hard to integrate. The major advantage of a generic approach is the guidance of the overall genre pattern during the whole analysis process, which helps to interlock detailed requirements. In plain words, the choice is between a requirements analysis ‘from user to document’ or ‘from document to user’. Each of them has its own specific advantages, and both can be used alternately or in parallel in an information engineering project. When a new website is set up and there is some uncertainty about its design, an initial survey of websites in the same business sector, in terms of genres, genre repertoire, and genre system, may be quite useful.
The discussion on ‘situatedness’ follows naturally from the foregoing argument. Genre knowledge is conceived as a form of situated cognition, i.e. our learning of genre occurs in situation bound activities (Berkenkotter & Huckin, 1995:7-13). For example, scientific writing is closely linked to activities in the laboratory and historical studies reflect the craftsmanship of analyzing historical records in an archive. The work of Orlikowski and Yates has demonstrated, how genre analysis may be used to acquire a better understanding of the users of information technology. However, information engineers may find the opposite direction even more interesting: genre analysis as a means of gathering user requirements in application design (e.g. Tyrväinen & Päivärinta, 1999). As mentioned above, genre analysis may be used to establish functional requirements in those circumstances where the users group is fuzzy and the actual tasks will vary considerably. Then the habitual use of documents may provide a better hold.
‘Situatedness’ is a more particular aspect of genre as a pattern of communication located in a discourse community. Taking this perspective may enforce a greater scrutiny in the functional design. User interface functions are sometimes introduced on logical grounds: a user needs a sort option because the list of items is quite long, a zoom function is required because pictures are initially displayed as thumbnails etc. There is nothing wrong with this line of reasoning, except it is superficial in a certain degree. Taking into account the reading process itself and the related social roles, as Paré and Smart suggested, we may better understand how these interface functions are precisely used in practice. In this way we get at a lower level of abstraction and generalization. Here the genre approach bears some methodological resemblance to the so-called ‘interpretive evaluation’ in usability engineering. This starts from the assumption that reality is an individual construction, and, therefore, focuses on the subjective understanding of reality and on complex interactions which may occur in the real world and that easily escape from the hypothesis-testing paradigm. A method like ethnography, for example, strives for a profound immersion in the situation to be studied, by analogy with anthropology, and places a strong emphasis on interpreting data in relation to the context [9].
Evolution. First of all, evolution occurs across time. Complementary to this, evolution is visible in the diversity of variants across contemporary discourse communities. In this respect digital genres are not fundamentally different from conventional genres. However, the Web is a single, universally accessible platform with numerous discourse communities all using the same stream of communication at once. Variants from different origin appear side by side, which makes counting instances of web genres not particularly easy. Moreover, discourse communities interact and genres get mixed, which may create some more confusion (Crowston & Williams, 1997). Differences between the survey by Crowston & Williams and that by Shepherd & Watters were partly explained by sample size and level of abstraction. For example, ‘catalogue’ and ‘resource’ (Shepherd and Watters) are broad genre categories, which cover more specific genres in the list of Crowston and Williams. Some documents were difficult to classify, because they seemed to be variants in a process of adaptation, or because the exact purpose was unclear.
Definition is a major problem. We may easily agree on the core of some well known genres (e.g. usual elements of a home page), but there will be always a large ‘gray zone’. Although defining genres may seem initially relatively unproblematic, it turns out to be a theoretical minefield in a more advanced stage. One theorist's genre may be another's subgenre or even supergenre. There are no rigid rules of inclusion and exclusion and of how to handle exceptions (Chandler). The right level of abstraction or the right point of view in setting up a classification might be always a matter of discussion: what may be right in one survey, may be not appropriate for an other.
Nevertheless, comparison is relevant, not only on a general level, i.e. how the Web is evolving. For competitive reasons we may want to analyze websites within a business sector, to find out how other companies have organized their information on the Web, in order to design a more distinctive website or, the opposite, to stay within a bandwidth of what is familiar and conventional. More refined techniques for measuring aspects of genre seem to be needed, which will require further research. A suggestion for another approach to this problem will be discussed below.
Transformation. Texts that are well structured, for example by means of a mark-up language as XML, are already actually broken up in components, which can be selected and reordered to new texts by means of scripts. In this way, the same data can be re-used for a variety of output documents. Conceptually, such a document is better considered as an “incarnation” of information components, which are stored in a database or textual repository, and happen to be displayed together as full text. Research on cybergenres has not yet grappled sufficiently with these new technological aspects. Crowston and Williams noted already that hyperlinks may modify genres. This is especially true for future forms of link behavior, not yet available in HTML, but already defined for XML. Links can be activated automatically on document display and the linked text may be embedded in the current text (transclusion), rather than it replaces the document that contains the hyperlink, as HTML- links operate nowadays. Such a transclusion may change form, content and function dramatically and will make generic classification even more difficult.
Moreover, it is far from clear how the phenomenon of information components is related to genre. We may hold on to a simple view and consider these components as elementary building blocks, which are genre-free. Then, genre applies only to the document itself, which is a superstructure, created on demand, with slots for information elements. Although this seems to be an attractive point of view, one can easily conceive of components that can hardly be designed without future functions (and forms) in mind: a document with an outline function must have components assigned to different levels, book references in a catalogue must have sort attributes, a clickable map must have a predefined granularity appropriate for functions on document level. So, the issue may be more complex and a lose coupling between components and genre seems to be more likely. All this is of practical importance for the creation and storage of content, but it goes also to the heart the matter: Can genre in this context still be reduced to a speech act? Or has modern technology disturbed this resemblance, because our production system of information has changed so profoundly and has been “de-humanized”? More research is required, here too, to answer these questions.
The review of literature on cybergenres makes clear, that, in addition to content and form, purpose and function are most relevant to modern genre analysis. As mentioned above, function in a digital context means also user interface functionality. Genres have been sometimes identified by careful coding of predefined features. Orlikowski and Yates distinguished purpose indicators (‘for your information’, proposal, question etc.), structural indicators (formatting features, headings, smileys etc.) and language indicators, such as informality, humor or sarcasm. A genre was defined inductively and based on a certain combination of features (Orlikowski & Yates, 1994). Dillon and Grushrowski (2000) followed a comparable strategy. In a Web survey with a great variety of genres and a relatively small number of instances less detailed coding may be sufficient. Then genres can be described conceptually on basis of characteristics of form and purpose (e.g. an essay is “a composition of moderate length on any particular subject” Crowston & Williams, 1997). Shepherd and Watters (1999) defined genres in their sample explicitly along the three dimensions of content, form and functionality (see table).
We noted already above, that these slight differences in method make comparison of results more difficult. Another flaw with counting at genre level only, is that specific combinations of content, form and functionality get lost. The use of animation, graphics or special formatting can be quite distinctive within a genre. While classified in the same category, the home page of a team park, for example, will be quite different from that of an undertaker (apart from content) as a result of different choices with regard to form and function. From the point of view of content engineering these intrageneric differences are interesting, because they represent different solutions for particular design problems. The intent of a team park's home page may be communicating movement, excitement and a feeling of an extraordinary world. The content engineering problem is, that this experience should be immediately apparent from the home page, since it is the symbolic index to all the team park has to offer. In generic terms: this particular content requires an appropriate combination of form and function, which is made visible in the choice of interface components. An animated menu with playful roll-over effects when the mouse cursor is moved over the items, may be a better choice than a static list of links. These combinations may be also used as indicators for measuring diversity and development in genres. Animation, especially with Flash and Shochwave, is quite popular, but has certain drawbacks. Such a genre analysis can provide practical justification for where animation is to be used and where it is not.
Content engineering may take some inspiration from software engineering. Digital documents with their embedded programming logic come close to software, or, to put it differently, electronic documents could be considered as programs with text as embedded data. Different programs of a kind show the same disparate variety in form and functionality as documents of a particular genre. Software engineering approaches this problem by classification on basis of so-called design patterns. “Each pattern describes a problem which occurs over and over again in our environment, and then describes the core of the solution to that problem, in such a way that you can use this solution a million times over, without ever doing it the same way twice” (Alexander et al., 1977). A design pattern names, abstracts, and identifies the key aspects of a common design structure that make it useful for reuse. It describes intent, problem, solution, and consequences in a conceptual manner, allowing the evaluation of design alternatives.
Recently, information reuse and link patterns in hypermedia have got more attention (Garzotto et al., 1996; Bernstein, 1998). Rossi, Schwabe and Garrido discussed the use of design patterns in hypermedia, in order “to motivate the hypertext community to discuss the problem and eventually produce a pattern catalogue in which hypermedia design experience is recorded as a set of related patterns..” (Rossi et al., 1997). They distinguished design patterns for hypermedia systems form those for hypermedia applications. The first category is more applicable to software development. The second includes navigational design and interface design and is more relevant to generic engineering problems. This is typically a field of interdisciplinary studies, where genre specialists meet the hypertext community.
An example of a navigational design pattern, relevant, for example, to the genre of Virtual Exhibitions, is ‘Navigational Contexts’. It deals with different collections of nodes (e.g. paintings, cities, persons, and objects that are part of multiple collections, such as Sun Flowers, which is part of Van Gogh's oeuvre and an item in the collection ‘Nature’). The use of each collection requires its own context, in terms of feedback, suggested links and the presentation of information. In ‘Navigational Contexts’ the navigational objects themselves are decoupled from the context in which they are to be explored. Contextual peculiarities are defined as ‘decorators’ that enrich the navigational interface when the object is visited in that context. (Rossi et al., 1997)
We have started the discussion above with two questions on the mutual relationship between genre theory and information technology. At first sight, the genre concept has not changed much by its application to digital communication. Looking back at a hundred years of theorizing, the first impression is, that emphasis has shifted from ‘author and work’ to ‘work and reader’, from production to reception. Within the triplet <content, form, function>, the last element has got much more attention with the advance of interactive digital documents. Generally speaking, however, the transition from printed documents to digital communication is more characterized by reference to the foundations of classical genre theory, rather than by conceptual innovations. But research on cybergenres is still relatively young and the innovative architecture of digital documents leaves some new questions to be answered.
The value of genre theory for content engineering is, that it helps to look at the design problems from a different perspective, and to ask a different type of questions with a broader scope than prevailing usability engineering strategies raise. Usually, application design follows the two lines of data modelling and functional modelling. The genre approach integrates both by drawing on some basic patterns of information handling we have learned as members of a discourse community. Its value could be best summarized as a form of ‘webliteracy’, but not in the ordinary, technical sense of the word, i.e. the basic skills of web surfing and writing HTML. Webliteracy in this context should also include social understanding and writing as a process of invention and composition, with electronic media and in a networked environment [10]. This will result in effective web presence and web identity, which then go beyond nice graphics and eye catching animations. The study of digital genres seems to be the primary key to this quality.
[1] Stanford Encyclopedia of Philosophy; D. Knowles, (1962). The evolution of medieval thought. London: Longman, ch.1.
[2] A good overview and critical comparison of genre theories (both Australian and American) is to be found in Knapp (1997). See also the introduction of Freedman & Medway, 1994: ch. 1.
[3] Refer to Knapp (1997:136) for a critical analysis of Miller's publications.
[4] The conference proceedings if HICSS are published by IEEE and available from the IEEE digital library.
[5] Both lists of bookmarks; the first type (hotlist) provides the list only, while the second (topical homepage) has some additional information concerning the link.
[6] Although these examples are taken from Crowston and Williams, Shepherd and Watters don't seem to follow these authors in their attempt to retrace the evolution of novel cybergenres to existing genres.
[7]A.J. Dix et al., (1998). Human-computer interaction. Prentice Hall, p. 268-279: Knowledge-based analysis and Entity-relationship-based techniques.
[8] See also: Good documents: http://www.gooddocuments.com/techniques/audience_m.htm, and An introduction to Web weaving (University of Minnesota): http://www.cee.umn.edu/dis/WW/.
[9] J. Preece et al., (1994). Human-computer interaction. Harlow etc.: Addison-Wesley, ch. 32.
[10] See also: J. December, (1996). A meditation on web literacy. CMC Magazine, October 1996: http://www.december.com/cmc/mag/1996/oct/last.html.
© Leen Breure, 2001