Home / Articles / Volume 6 (2009) / Thematic and situational features of Chinese BBS texts
Document Actions



Linguistic features have long been regarded as marking particular components of the communicative situation. In systemic functional grammar (SFG) (Halliday, 1994/2000; Halliday & Matthiessen, 2004), texts are viewed as social processes and the context of a text is manifested through language, especially on the level of lexicogrammar. The SFG framework has been applied to investigate various kinds of texts or discourse, but it has rarely been applied to computer-mediated discourse, despite the fact that the latter is “highly sensitive to a variety of technical and situational factors” and that aspects of both technical and social context “influence discourse usage in CMC environments” (Herring, 2007, n.p.). How elements of context have influenced the linguistic features of computer-mediated discourse and what particular discourse usages or linguistic features are illuminated through SFG remain significant unaddressed problems.

Many linguistic studies of computer-mediated communication (CMC) or computer-mediated discourse (CMD) (Herring, 1996, 2001, 2007) have focused on descriptive accounts of features such as emoticons, unconventional spellings, writing systems, sentence length, and lexical characteristics (e.g., Al-Sa’di, 2005; Anis, 2007; Nishimura, 2003; Palfreyman & Al Kbalil, 2007; Su, 2007; Yang, 2007). There are also numerous studies concerning the interactive nature, media constraints, participant roles, relationships among participants, and purposes and functions of CMC (e.g., Chung, 2008; de Oliveira, 2007; Georgakopoulou, 2004; Goutsos, 2005; Jones, 2004; Yates, 2000).

Until recently, little research had been done on CMC in languages other than English. Recently, however, “researchers have turned their attention to other languages on the Internet, often their native languages” (Danet & Herring, 2007, p. 5). This includes a growing body of literature in the investigation of Chinese CMC.1 Most of these studies have focused on the adaption of the Chinese writing system in Internet communication, or on linguistic or discourse features of CMC in Chinese. For example, Yang (2007) identified five types of adaption of the writing system in Internet language in mainland China, such as stylized Mandarin, stylized dialect-accented Mandarin, and stylized English. Su (2007) also investigated the creative use of writing systems on the bulletin board systems (BBS) of two college student organizations in Taiwan. Gao (2006) investigated the influence of English on Chinese linguistic features based on data from BBS, chat rooms, Internet novels, personal emails, and other posts at public websites in mainland China, and Lee (2007) investigated the mixture of Cantonese and English elements, as well as morpheme-by-morpheme literal translations, by collecting data from ICQ messages of undergraduate students in Hong Kong. Wang, Katz, and Chen (2003) narrowed their investigation to a specific Chinese word shuo (‘say,’ 说) and investigated its grammaticalization based on data from BBS talk and Mandarin conversations in Taiwan. Other research has focused on the communicative situations, such as participant roles, relationships among participants, and the purposes and functions of CMC in Chinese. Gao and Yuan (2005) discussed the construction of modernity through language use in CMC in mainland China. Xie (2008), focusing on the construction of relations among older Chinese Internet users, analyzed the multidimensional relationships that developed through CMC.

The present study takes a different perspective by adopting the framework of SFG to investigate: 1) thematic features of Chinese texts selected from an electronic BBS in mainland China; and 2) the relations between thematic features and situational features. As a concept derived from the Prague School and later adopted in the SFG framework, theme “serves as the point of departure of the message; it is that which locates and orients the clause within its context” (Halliday & Matthiessen, 2004, p. 64). A closely related concept is rheme, the remainder of the clause and the part in which theme is developed. The thematic status of an item is signaled by its initial position in the clause. Thematic features in the present study are identified in terms of type of theme—simple, multiple, or clausal theme; type of multiple theme; nominal groups and pronouns in the thematic position and personal pronouns in the thematic position.

The SFG framework was adopted because it was constructed for the purpose of text analysis (Fang, 2005; G. W. Huang, 2005) in order to “make it possible to say sensible and useful things about any text” (Halliday, 1994/2000, p. xv). Text is “the most popular form of CMC in use today, even in graphical chat environments” (Danet & Herring, 2007, p. 6). The Chinese BBS in the present study is no exception—that is, messages posted by its participants are mainly text-based, with a low incidence of uploaded pictures or videos.2 Because a comprehensive application of SFG to CMC texts is beyond the scope of a single article, the present study focuses on one aspect of SFG, i.e., thematic features, and investigates their relationship to situational features. Specifically, the study addresses the following questions:

  1. What are the thematic features of Chinese BBS texts?

  2. What is the relationship between the thematic features and situational features of the BBS? Situational features in this study include both medium factors (Herring, 2007) and situational (social context) factors (Biber, 1988; Herring, 2007).

  3. Are the thematic features determined primarily by mode, one of the three elements of context according to the SFG framework (Halliday, 1978/2001; Halliday & Hasan, 1985) along with field and tenor? According to Halliday (1978/2001), thematic features of a text are determined by its mode, viz. the distinction of medium and “the symbolic forms taken by the interaction” (p. 144).

Theoretical Background

In addition to the theme-rheme structure of a clause, SFG has established a relationship between lexicogrammar, of which theme is one of the constituents on the clausal level, and elements of context. The three elements of context are field, tenor, and mode (Halliday, 1978/2001; Halliday & Hasan, 1985). Field refers to the nature of the activity; tenor refers to the role relationships in the situation and speech roles that come into being through the exchange of verbal meanings; and mode refers to the symbolic forms taken by the participants, the distinction of medium, and extends to the semiotic functions the text is serving in the environment. According to Halliday (1978/2001), the three elements of context determine the semantic patterns of a text, which are realized by lexicogrammatical particulars. What concerns the present study is the last element: When a speaker encodes “the feature of the channel, the rhetorical mode and so on (the mode), he draws on the textual component” (Halliday, 1978/2001, p. 63), which is embodied in thematic features of a text. In other words, as a realization of textual function, thematic features are determined by the mode. Based on the SFG framework, therefore, we can draw the inference that thematic features of CMD are determined by the networked computer or other medium factors of CMC.

In addition to Halliday and Hasan’s (1985) tripartite division of context, numerous classifications of context exist (e.g., Brown & Fraser, 1979; Duranti, 1985; Fishman, 1972; Hymes, 1974; Renkema, 2004). Despite their disagreements, these classifications share three basic dimensions of context that fit with Halliday and Hasan’s tripartite division of context: the physical dimension, the dimension of the activity being engaged in, and the dimension of participants (Jones, 2004; Murray, 1988). Based on Halliday and Hasan’s (1985) tripartite division into field, tenor, and mode, the present study draws on Biber (1988) and Herring (2007) for an elaboration of situational features with respect to their relations to thematic features.

Biber’s taxonomy of situational features comprises eight categories, namely: participant roles and characteristics, relations among the participants, setting, topic, purpose, social evaluation, relations of participants to the texts, and channel, each of which includes its respective sub-components. Herring (2007) has provided a faceted classification scheme specifically for CMD, which includes both technical and social factors that potentially influence discourse usage in CMC environments. Under each of the two categories, a number of facets are posited based on empirical evidence from the CMD research literature. The category of medium (or technical) factors includes facets such as: synchronicity of participation, persistence of transcript, anonymous messaging, and quoting. The category of situational (or social) factors includes: participation structure, participant characteristics, purpose, topic, tone, activity, norms, and code.

The Identification of Theme in Chinese

Although the concept of theme was put forward on the basis of English, Halliday (1994/2000) has provided a general guide to the identification of theme in other languages:

If in any given language the message is organized as a theme-rheme structure, and if the structure is expressed by the sequence in which the elements occur in the clause, then it seems natural that the position for the theme should be at the beginning, rather than at the end or at some other specific point. (p. 38)

Chinese is a language in which utterances are organized in a theme-rheme structure, and the framework of SFG has been applied to the analysis of Chinese by several researchers (e.g., Fang, 1989; Fang & McDonald, 2001; X. J. Hu, 1997; Z. L. Hu, 1994; Li, 2007). Theme in Chinese is generally identified as the left-most constituent of a clause, and is further classified into simple theme, clausal theme, and multiple theme. A simple theme is one that “consists of just one structural element, and that element is represented by just one unit—one nominal group, adverbial group or prepositional phrase” (Halliday, 1994/2000, p. 39). Clausal themes appear in clause complexes in which the clause in the initial position of the clause complex is identified as a clausal theme. Both clausal themes and rhemes may further contain their respective themes and rhemes.

Elements of multiple themes may include topical themes, interpersonal themes, and textual themes.3 A topical theme is used to construe human experience, either a process, participants in the process, or any circumstantial factors; a textual theme is used to indicate the relationship of a clause with its preceding or following clauses and includes discourse signalers like “yes, no, well, oh,” conjunctions like “and, when, until, even if, in spite of the fact that,” and conjunctives like “in other words, moreover, meanwhile, in this respect;” and an interpersonal theme shows the opinion or attitude held by speakers and includes vocatives, i.e., personal names used to address, and modal adjuncts such as “probably, in my opinion, at first, surprisingly.” With topical theme as an obligatory element, a multiple theme can take the form ‘tex+top,’ ‘inter+top,’ or ‘tex+inter+top.’ (1) is a clause taken from the data set that contains a multiple theme with all three elements: textual, interpersonal, and topical themes.

(1)4 Thread 5: The tragic and subversive features of Leslie Cheung (a Chinese movie star)

不过,         实话,         《家》、《春》、         《秋》时      感情5
buguo,   shuo  jü   shihua,     du     jia          chun       qiu         shi     de    ganqing   
but,        tell     CL truth,       read  Family  spring    autumn time   MM empathy   

不如          读     《平凡     的     世界》  来得   强烈……
buru           du     pingfan de shijie  laide qianglie
unequal-to  read  Ordinary  MM  world    come  strong

However, to tell the truth, my empathy for The family, spring, and autumn is not so strong as that for The ordinary world.’6

In the above example, the conjunction buguo (‘but’) is a textual theme, expressing the relationship of the clause with previous clauses; the modal adjunct shuo jü shihua (‘to tell the truth’) is an interpersonal theme used to express the speaker’s attitude; and du jia chun qiu shi de ganqing (‘my empathy for The family, spring, and autumn’) is a topical theme used to construe the speaker’s experience.

General Internet Usage in China

Before presenting a detailed description of the Chinese BBS from which the data for the present study were selected, we shall first take a look at general Internet usage in mainland China. According to The 21st China Internet Research Report, a report of a survey of Internet usage in China conducted by the China Internet Network Information Center7 (CINIC) December 8-31, 2007, Internet usage in China is developing where individuals can participate actively on the Internet as broadcasters, writers, and producers of Internet content. As computer systems that serve both as information sources and as forums for particular interest groups, BBSs are a major way for Internet users to participate actively on the Internet. The CINIC report shows that about two-thirds (65.7%) of the 210 million Internet users in China had contributed posts or uploaded texts, pictures, or videos to the Internet. Most of the content provided by Internet users in China is text based (35.4%).

Data and Methodology

Kaidi BBS

The data in the present study come from messages sent to a BBS hosted in mainland China called Kaidi.8 With more than 5.64 million registered members,9 the Kaidi BBS is popular among Chinese Internet users for its comprehensive coverage of social phenomena and in-depth comments contributed by members. It has been listed as a famous BBS among Chinese Internet users, along with other well-known BBSs such as sina BBS,10 sohu BBS,11 and tianya club.12

Kaidi BBS is divided into 22 discussion groups, each of which focuses on a topic, such as economy, travel, culture, sports, health, law, social phenomena, and entertainment. Under each topic, registered members can start new discussions or join existing discussions to express their opinions. A title is required when starting a new discussion, and all the titles are listed on index pages. A click to a specific title enables readers to trace all the posts under the title; these are collectively known as a thread.


Six threads13 were selected from the discussion group discussion on entertainment14 on the Kaidi BBS. The posting dates of the six threads range from November 2004 to August 2006. Although discussions in the group focus on entertainment, they differ from typical fan sites, which mainly attract young people and highly involved fans (cf. Nishimura, 2003). In the discussion of entertainment in the present study, most participants take an objective position and provide thoughtful reflections or rational comments on films and/or stars.

The six threads were selected to guarantee a certain degree of similarity and comparability. First, they were selected from the same discussion group to ensure that they shared the general topic of commentary on films, TV series, and/or stars. Second, each of the six threads has its own explicit discussion topic. Although new discussions on the Kaidi BBS are started from time to time, not all of them possess the prestige and correctness of traditional publicly-available written texts. Some posts contain generalities or gossip, or are irrelevant advertisements. This is not the case in the six threads, all of which have an explicit discussion topic, as shown in Table 1. Finally, the six threads were selected to guarantee a certain amount of interaction among participants. The average number of messages (or posts) in each thread is around 33; they were posted by multiple participants, with an average of 22 participants per thread. This makes it possible to generalize the findings about thematic features to a range of participants, instead of the findings being limited to the personal preferences of individual participants.

Table 1. Detailed information about the six threads

Within each thread, a distinction is made between initial posts and following posts. On the one hand, initial posts of the six threads are all pre-edited and contributed by the participants who initiate the thread. They are usually the first post of the thread; other registered members of the BBS can post messages subsequently if they are interested in the initial post or the discussion topic. Contributors of initial posts are unable to predict who will take part in the following discussions, hence there are no cross references to the following participants or their posts in initial posts. Initial posts in the six threads are much longer than any following posts and can be taken out of the thread to stand alone, as is also indicated by the use of titles or even subtitles in many initial posts. On the other hand, posts following the initial post can extend the discussion, provide background information, or make a compliment to or abuse the initial post or its contributor. Therefore, the total number of Chinese characters, T-units (cf. the ‘Methods’ section), and themes were counted in the two kinds of posts separately (see Table 2).

Table 2. T-units, number of Chinese characters, and theme total in initial and following posts
Notes: 1) ‘Theme total’ includes simple themes, multiple themes, and clausal themes, as well as themes in the theme/rheme part of clausal complexes; 2) Initial posts are referred to as ‘initial’ and all following posts as ‘following’ in the tables and figures.

There are 1,211 themes and 739 T-units in the initial posts, and 1,489 themes and 1,076 T-units in the posts that followed. In total, there are 41,874 Chinese characters, with 20,472 in the initial posts and 21,402 in the following posts. The distinction between initial posts and following posts was also applied to the investigation of different thematic features in the six threads.


A frequency count was carried out to examine the thematic choices in the declarative clauses of the six threads. The count of thematic choices was restricted to independent clauses, “those that can stand by themselves as a complete sentence” (Halliday, 1994/2000, p. 43). Minor clauses were excluded from the count, because they are made up of single words that realize speech functions such as exclamations, greetings, or alarm, e.g., ding (‘I support the topic’), tong’gan tong'gan (‘Share your feeling’), daixü… (‘To be continued…’), hao! (‘Good!’). Some messages (or part of them) were excluded from the count: messages (or part of them) quoting from previous posts or copying from other sources, messages deleted by the site manager, redundant messages contributed by the same participant in the same thread, and uploaded pictures. Of the 221 messages posted, 29 were excluded from the count.

The unit of analysis of thematic features in the present study is the T-unit (Fries, 1995), which is an independent clause together with all the clauses that are dependent on it. This approach avoids fuzziness in the identification of clauses or clause complexes (Thompson, 1999). It is applicable to the analysis of clauses in Chinese texts, especially when the random use of punctuation and lack of formal constraints on Chinese clauses make it difficult to establish clear-cut boundaries between clauses.

Findings and Discussion

The investigation of thematic features in the present study focuses on: 1) the choice of simple theme, multiple theme, and clausal theme; 2) the choice of the three types of multiple theme; 3) the distribution of nominal groups and pronouns in the thematic position; and 4) the distribution of personal pronouns in the thematic position. The findings for the thematic features are interpreted in relation to the situational features of the BBS.

Choice of Simple Theme, Multiple Theme, and Clausal Theme

The majority of all themes are simple themes (around 54% in both kinds of posts), and the percentage of clausal themes in initial posts is higher than in following posts.

Figure 1. Distribution of the three types of theme

The higher proportion of simple theme than clausal theme is related to the purposeandparticipant roles and characteristics of the BBS. Purpose refers to the “outcomes that participants hope for, expect, or intend from the communication event” (Biber, 1988, p. 32), or the goals of interaction of individual participants (Herring, 2007). With the purpose of engaging other participants in discussion (Collot & Belmore, 1996), authors of BBS texts tend to try to make their texts accessible to readers instead of being obscure, which may lead them to prefer simple themes. In contrast with multiple themes and clausal themes, simple themes are characterized by a single structural element represented by just one unit. Clausal themes require more effort for comprehension, due to the complicated sentence structures of clause complexes. Thus clausal themes are less preferred than the single structural element of simple themes. The choice of simple theme avoids grammatical complexity and makes texts more reader friendly.

Another component of the situational features—participant roles and characteristics (Biber, 1988; Herring, 2007)—may also play a role in shaping the choice of simple themes. BBS participants are assuming the tripartite roles of addressors, addressees, and recipients. When composing their messages, participants would bear in mind they are addressing a large anonymous public rather than writing to intimate friends or themselves. Awareness of addressing the public and a desire to get positive responses may lead to a preference for simple themes. This is in accordance with Collot and Bellmore’s (1996) claim that BBS texts share similarities with public interviews, in which “the interviewees respond specially to the interviewer but do so for the benefit of a wider audience” (p. 26).

The comparatively higher proportion of clausal theme in initial posts can be accounted for by the different purposes of the two kinds of posts, with writers of initial posts posting to explicate their points and those posting after the initial posts responding to previous posts or interacting with other participants. Initial posts do not take part in the overall interaction directly except by providing a discussion topic for the following posts. When participants compose a message to start a new thread, they are unable to predict what kind of interaction will come after or who will participate in the following discussion, or they may choose to ignore such factors because they are not writing exclusively to meet the requirements of potential recipients. As a consequence, they might not avoid using clausal themes when necessary, even if it runs the risk of requiring more comprehension effort from their readers. In contrast, those who post afterwards tend not to compose fully-developed messages like initial posts; otherwise they could start new threads by themselves. These participants contribute to the thread mainly to make comments on the initial posts or the discussion topic, or to respond to certain participants who posted before them, which requires them to take into consideration the accessibility of their posts. The different purposes of the two kinds of posts are reflected in their different usage of simple themes and clausal themes.

Choice of the Three Types of Multiple Theme

The above discussion did not cover multiple themes, because an analysis of multiple themes needs to clarify its constituting elements first. With topical theme as an obligatory element, a multiple theme is possible in the form of ‘textual theme + topical theme’ (tex+top), ‘interpersonal theme + topical theme’ (inter+top), or ‘textual theme + interpersonal theme + topical theme’ (tex+inter+top). Textual themes can be composed of discourse signalers, conjunctions, or conjunctives; interpersonal themes can be composed of vocatives or modal themes.

Figure 2. Distribution of the three types of multiple theme

Figure 2 shows that the first two types of multiple theme, ‘inter+top’ and ‘tex+top,’ constitute the largest proportions, while the third type accounts for a low proportion in both kinds of posts. This finding is consistent with Halliday’s (1994/2000) claim that the first two types are “regularly found in most types of discourse” (p. 54), while the last type is comparatively rare. In addition, the percentage of ‘tex+top’ is much higher than ‘inter+top’ in both kinds of posts.

The higher frequency of ‘tex+top’ than ‘inter+top’ may be partially attributed to the naturally thematic position of conjunctions. As one constituting element of textual theme, conjunctions are inherently thematic: “[I]f they are present in the clause at all, they come at the beginning” (Halliday & Matthiessen, 2004, p. 83). However, modal adjuncts, one of the constituting elements of interpersonal theme, have a more flexible position in the clause; the speaker or writer can choose whether to put them in the thematic position or not (Thompson, 1999).

In addition to the naturally thematic properties of conjunctions, situational features like thepurpose for requesting and giving information, the specified topic, relations among participants, and participant roles and characteristics, as well as one of the medium factorsof the BBS, may also play a role in shaping the distribution of multiple themes in the present data.

The high level of ‘tex+top’ may be attributed to the purpose of the BBS, giving or requesting information and engaging in discussion. Textual themes could be used to signal logical links within the text and could help to accomplish the goal of giving information by enhancing the logical and cohesive power of the texts. Textual themes such as “but,” “therefore,” and “in addition” are frequently used in the BBS messages, which are a kind of persuasive text (Collot & Belmore, 1996), especially in initial posts, to signal logical and argumentative links, as in the following example:

(2) Thread 1: Where goes the future of Hong Kong movies

但是作为一个新锐导演其某些想法还是得到了业内的肯定,//15 而且投资方也对他抱有更多的期望,// 但是却继续沉迷自己所谓的黑帮文化和mtv情结……

but as a vanguard director, some of his ideas have been accepted by the professionals in the film industry. // In addition, the investors have high expectations of him. // However, he continues to be obsessed by his so-called sinister gang culture and fascinated by some techniques of MTV.

In the above example, a series of textual themes—“but,” “however,” and “in addition”—are used to demonstrate the logical and argumentative links between successive clauses.

Textual themes are also used dynamically and interpersonally to signal transitions in the interactive development of texts. The high frequency of textual themes in spoken Chinese (Li, 2006) suggests that it is not really an effort for hearers to decode utterances with textual themes, and that such themes do not hinder the smooth progression of conversations. With the dual function of helping to give information and engage in the discussion, textual themes with compulsory topical themes form the type of multiple themes, i.e., ‘tex+top,’ that constitute the highest proportion in both initial and following posts.

The higher proportion of ‘tex+top’ in initial posts than in following posts (76% to 59%) and the higher proportion of ‘inter+top’ in following posts than in initial posts (28% to 17%) can be attributed to the different purposes of the two kinds of posts. Textual themes are preferred by contributors of initial posts to signal logical links within their messages, and interpersonal themes are preferred by contributors of following posts to enact social relationships among participants. However, few contributors of initial posts would identify their primary purpose for posting as being to enact any social relationship with potential recipients. Therefore interpersonal themes, including personal names, are comparatively rare in initial posts.

The generally higher proportion of ‘tex+top’ than ‘inter+top’ may also reflect the fact that the purpose of requesting and giving information outranks the purpose of maintaining harmonious relationships among the participants. Requesting and giving information is closely related to the topic of the discussion group. With prescribed and specified topics, posts on the BBS are more likely to convey some propositional information than to establish or reaffirm interpersonal relationships among participants, which is frequently realized by interpersonal themes. This results in a lower proportion of interpersonal themes than textual themes in the BBS messages.

Relations among participants (Biber, 1988) or participation structure (Herring, 2007) of the BBS may also contribute to the low proportion of interpersonal themes. Most participants in the BBS take part in the discussion using nicknames or pseudonyms; they are unknown to each other in real life, and rapport among them can only be established gradually through discussions or interactions on the BBS. Moreover, BBSs usually lack a stable membership, long-term commitment, and social accountability (Jones, 1995, 1998). Few participants consistently join in the discussion, let alone develop long-term relationships with others. Given the difficulty of establishing and maintaining stable personal relations, participants in the BBS may be less likely to use interpersonal themes, which are mainly used to enact social relationships.

Another situational feature, participant role and characteristics (Biber, 1988; Herring, 2007), may also influence the use of interpersonal themes. One of the types of interpersonal theme is vocatives, which include names and other forms of direct address. When composing their messages, BBS participants, especially those who post initial posts, are aware that they are addressing a large anonymous public rather than a clearly identified recipient. Lack of specific addressees makes vocatives less likely to appear in the thematic position to form any interpersonal themes.16

In addition, the low proportion of vocative in the thematic position may be related to one of the medium factors of the BBS, the availability of the strategy of quoting (Herring, 2007). When replying to others, participants of the BBS have several devices available besides using vocatives. For instance, they can click ‘reply’ (the option is provided in the heading of each post), and the BBS system will automatically include the responded-to text, together with the information who posted the responded-to text and the time it was posted, which reduces the need to use vocatives in the response. Or participants can use other strategies, as illustrated in the following example:

(3) Thread 6: Pure love is almost impossible for a woman – a review of the movie ‘King Kong’

to     勒追      和     恶意教唆:
to     Lezhui   he      Eyijiaosuo:
to     Lezhui   and    Eyijiaosuo (two participants in the discussion):

其实,       真的      还    不    懂                究竟       什么         是    爱情。
qishi,       wo   zhende   hai   bu    dong            jiu’jing     shen’me    shi    ai’qing.
actually,   I      really      yet   NG  understand   exactly     what         be    love.

‘To Lezhui and Eyijiaosuo:
Actually, I don’t know what exactly love is.’

When replying to two other participants, the writer used a combination of English and Chinese to address them, i.e., to 勒追和恶意教唆 (‘to Lezhui and Eyijiaosuo’), which avoids putting vocatives in the thematic position of the clause.

Distribution of Nominal Groups and Pronouns in Thematic Position

This section compares the distribution of nominal groups and pronouns in the thematic position. Pronouns in this section include personal, indefinite and demonstrative pronouns. Figure 3 shows that the percentage of nominal groups is higher than that of pronouns in both kinds of posts. The lower percentage of pronouns may be partly attributed to one of the typological features of Chinese, widespread zero anaphora. “The co-referential entities to the topic often occur in zero form, whereas in English they normally are overtly coded in pronouns” (Wu, 2004, p. 11), especially when the identity of referents “can be easily established through linguistic and extralinguistic information available in the discourse” (Chen, 1986, p. 194).

Figure 3. Distribution of nominal groups and pronouns

The distribution of nominal groups and pronouns is also related to the following situational features: a lack of shared temporal and physical settings, specified topics, the purpose of giving and requesting information, and relations of participants to the text, i.e., the strategies of planning and editing available to the participants.

Lack of shared temporal and physical settings in the BBS calls for specific references, which may contribute to the high proportion of nominal groups in the thematic position. Setting refers to different aspects of the physical and temporal context, such as when and where the communication takes place; most important is the extent to which participants share time and space (Biber, 1988). As an asynchronous communication tool, BBSs do not “require that users be logged on at the same time in order to send and receive messages” (Herring, 2007, n.p.). Users can access or reply to posts at their own convenience, and their posts do not necessarily require an immediate response. They come into contact with each other only through networked computers, and messages posted on the BBS are the only thing they share. They have to rely primarily on text to convey their opinions, and communications on the BBS do not receive any assistance from paralinguistic features and prosodic information, such as intonation, stress, or gestures. The lack of visual or audio contact and immediate feedback means that participants cannot rely on shared context to make their meanings—for instance, the referents of pronouns—clear (Crystal, 2001).

The clearly identified general topic of the discussion forum, i.e., entertainment, and the explicit sub-topics of the six threads may also contribute to the high level of nominal groups in the thematic position. As pointed out in The guide to the discussion on entertainment,17 which articulates the norms of posting in the group, the group’s discussion topics are clearly delimited. Moreover, the six threads were selected because they focused on explicit discussion topics and generated significant discussions. Therefore, posts in the discussion group, at least in these six threads, are mainly used to convey ideational functions (Biber, 1988), i.e., propositional and informational content, rather than interactional functions, as in the usual case in everyday conversation. Compared with pronouns, nominal groups carry more propositional and informational content, hence are more likely to accomplish the goal of conveying ideational functions.

Furthermore, the high proportion of nominal groups in the thematic position could also be related to the purposeof the BBS. In the present study, the purpose of requesting or giving information outranks the purpose of maintaining harmonious relations among the participants; this seems to play a role in shaping the BBS texts as a “non-narrative and highly persuasive discourse type” (Collot & Bellmore, 1996, p. 26). Contributors, especially of initial posts, tend to provide informational content when composing their posts, with the goal of convincing other participants of their opinions; this may lead to a preference for nominal groups in the thematic position, which is reflected in the slightly higher proportion of nominal groups in initial posts than in following posts (see Figure 3).

Finally, the relation of the participants to the text (Biber, 1988) may also influence the distribution of nominal groups and pronouns in the thematic position. The present study has made a distinction between initial posts and following posts. Initial posts, which are comparatively long and well organized, are presumably planned beforehand and pre-edited. Messages could also be edited before being posted with word processing software. However, the strategy of editing and planning is not confined to initial posts. Tags, which appear at the end of some posts and say that “This post has been edited by the author at XX: XX (specific time),” give an explicit indication that some posts have been edited after being posted for the first time. Strategies of editing limit the number of unplanned, spontaneous texts that feature fragmentary structures and unconventional word orders. When editing their texts, BBS participants would presumably remove or replace pronouns if ambiguities could possibly arise.

Distribution of First, Second, and Third Person Pronouns

Within personal pronouns, a further distinction was made between first, second, and third person pronouns in initial posts and following posts. As shown in Figure 4, the general distribution of the three kinds of personal pronouns almost runs in parallel in the two kinds of posts, with a bottom-up ranking of their proportions from second person pronouns, third person pronouns to first person pronouns. The distribution of the three kinds of personal pronouns corresponds to their different degrees of dependence on the setting.

Figure 4. Distribution of personal pronouns

In terms of the dependence of first person pronouns on the setting, the use of first person pronoun is less constrained by the physical and temporal setting, as compared with second and third person pronouns. The intended referent of a first person pronoun is usually the person who composed the message, regardless of the physical and temporal settings. In contrast, the interpretation of a second person pronoun (whether it is used to refer to potential readers or participants in the discussion, and exactly who the participant being referred to is) is highly dependent on the physical and temporal settings. Consequently, the lack of shared setting leads to the low proportion of second person pronouns in both kinds of posts (11% and 7%). Moreover, even when there is a possibility of using second person pronouns, nominal groups are used instead to avoid ambiguity. Third person pronoun, whose intended referent can be traced from the surrounding text, takes the intermediate ground (see Figure 4) by neither relying exclusively on physical and temporal setting for interpretation nor having a participant role as its default referent.

First person pronouns have taken the highest proportions among the three kinds of personal pronouns in both kinds of posts. Although there is no remarkable difference between the average percentages of first person pronouns in both kinds of posts (49% and 51%), the distributions of first person pronouns in the six threads differ in initial posts and following posts. The percentage of first person pronouns among the three kinds of personal pronouns in initial posts ranges from 0% to 100%, while the percentage of first person pronouns in the following posts is relatively consistent, ranging from 36% to 67% (see Table 3).

Table 3. Distribution of first person pronouns

The further investigation of the distributions of first person pronouns in the six threads may reflect the different purposes of the two kinds of posts. As discussed above, participants post after the initial posts mainly to make comments on the initial posts or discussion topics or to respond to certain posts or participants, whereas writers of initial posts typically seek to explicate their points in well-developed, organized messages. The use of first person pronouns in initial posts is highly influenced by the writing styles of different contributors. In the extreme case of thread two and thread six, no first person pronoun is found in the initial posts. Furthermore, the lack of shared setting may also lead to the necessity to present oneself from time to time (Yates, 1996), especially in the case of following posts whose major purpose is to express opinions from different points of view. Therefore when participants contribute after initial posts, they tend to start from their point of view to express different opinions or make comments, using structures like ‘I think…,’ ‘I feel like…,’ or ‘I (once did)….’

A Functional Approach to the Generalization of Situational Features and its Problems

The relationships between situational features and thematic features discussed in previous sections are summarized in Table 4. Situational features in the present study, including participants’ roles and characteristics, relations among participants, setting, topic, purpose, relations of participants to BBS texts and media factors, are further generalized into the three components of context: field, tenor, and mode (Halliday, 1978/2000; Halliday & Hassan, 1985). Tenor (the role relationships in the situation) includes participant roles and characteristics, and relations among participants; field (the nature of the activity) includes setting, topic, and purpose; and mode (the symbolic forms taken by the participants) includes relations of participants to the text and medium factors.

Table 4. Relations between thematic types and situational features

According to Table 4, topic, purpose, and setting—the components of field—are most important in determining thematic types. In other words, the thematic features found in the present study tend to co-occur with field more often than with mode, which seems to go against Halliday’s claim that thematic features are determined by mode. As mentioned above, SFG distinguishes three strata of language, i.e., lexicogrammar, semantics, and situation (or context). Table 5 shows the relations among the three strata of language in the SFG framework.

Table 5. Relations among the three strata of language (after Halliday, 1978/2000; Halliday & Hasan, 1985)

As the last column of Table 5 shows, the SFG framework presupposes that mode determines the textual meaning of a text, and thematic features in turn realize the distinctive features of the mode. However, mode does not appear to play a significant role in shaping thematic features in the present study. Why does this discrepancy arise?

The present study explores the discrepancy from three angles. First, there is an overlap between the definitional scopes of field and mode. Field refers to the nature of activities. When the social activity is inherently of a verbal nature, the verbal interaction among the participants is the whole of this social action, or in Yates’ (1996) words, “there is no such field beyond the focus of the interaction” (p. 45). The setting, topic, and purpose, which describe what is going on in the verbal interaction, are counted as components of the field. In contrast, mode refers to the symbolic forms taken by the interaction and the distinction of medium, e.g., spoken, written, or computer-mediated. However, Halliday (1978/2001) has extended mode to cover the functions that texts serve in the total situation, which makes mode too inclusive a term to draw a clear line between field and mode. In the case of verbal interaction, the function served by the texts overlaps with the purpose accomplished by the texts. As is shown in the present study, the purpose of the discussion is to convey propositional information and interact with other participants, which can also be said to be functions of the texts (see Figure 5).

Figure 5. Overlap between the definitional scopes of field and mode

Second, the realization of textual meaning may not be confined to thematic features. As a clause-internal aspect of texture (Halliday, 1978/2001), theme on the clause level has been found to be insufficient to account for the structure of a text (Hu, Zhu, Zhang, & Li, 2005). Halliday (1994/2000) claims that the theme-rheme structure realizes the textual meaning of a text, i.e., how a message fits in with the other messages around it and with the wider context of interaction. However, the issue of theme-rheme structure has not been addressed by Halliday or in the SFG framework. The pattern of theme-rheme structure development in texts, i.e., thematic progression patterns, which was first put forward by Danes (1964), has long been suggested and applied in practical research as a complement to the investigation of thematic features (e.g., Hu et al., 2005; Y. Huang, 1985; Liu, 2006; Nwogy & Bloor, 1991).

The last point resides in the appropriateness of listing mode, or more specifically, the distinction of medium and rhetorical mode, as a component having equal status with field and tenor. Does medium play so important a role in determining the choice of themes in clauses? For instance, the first person pronoun in the thematic position is identified by Halliday (1994/2000) as a mark of informal conversation, but further inquiry into its relation with informal conversation may lead to an exploration of field and tenor. The position of mode in relation to field and tenor has been addressed by Martin (1992/2004), who separated mode from the listing with field and tenor, and identified it as “a resource for constructing interaction” (p. 514). If Martin’s perspective is taken, mode becomes a cover term for field and tenor, in which case genre may be a more appropriate term than mode. In addition to the problematic position of mode, the bi-unique relations between lexico-grammatical resources and context (Halliday, 1978/2001; Halliday & Hasan, 1985) have been questioned and discussed by several researchers (e.g., Hasan, 1995; Thetela, 1997, cited in Zhu, 2005; Xu, 2004; Zhu, 2005), although a generally acceptable framework is still lacking.


This study investigated thematic features of Chinese BBS texts and discussed their relations with situational features of the BBS. It found a preference for simple themes as opposed to multiple themes or clausal themes. It also found that interpersonal themes occur less often than textual themes; nominal groups are preferred to pronouns in the thematic position; and first person pronouns are more likely to occur in the thematic position than second and third person pronouns. Situational features (Biber, 1988; Herring, 2007), found to co-occur with thematic features include: participant roles and characteristics, relations among participants, setting, topic, purpose, relations of participants to BBS texts and media factors, which were further generalized into Halliday’s tripartite classification of context, namely, field, tenor, and mode.

According to the SFG framework, theme is conceptualized as a realization of the textual meaning, which is in turn determined by mode. However, the present empirical study of Chinese BBS texts finds that thematic features are not determined by mode alone; thus, the bi-unique correlation between thematic features and mode seems to be untenable. Thematic features of the BBS texts are related to all three components of context, especially to the purpose of the BBS, which is a component of field. The mode of the BBS—which includes medium factors and relations of participants to the text—plays a role in shaping thematic features of the texts, but it is not a determining role, nor the only role. The present study thus opens up a discussion about SFG, particularly the relationship between thematic features and situational features, in the context of CMC and the Chinese language.


1. Chinese characters differ from the Roman alphabet, and keyboard entry of Chinese characters has used various software programs based on either the shape or pronunciation of Chinese characters. See Su (2007) for a detailed description of the inputting of Chinese words.

2. Based on The 21st China Internet Research Report provided by the China Internet Network Information Center (CINIC). See also the section in this article titled ‘General Internet Usage in China.’

3. These are abbreviated respectively as ‘top,’ ‘inter,’ and ‘tex’ hereafter.

4. Chinese examples in single clauses are written in the pinyin system of romanisation. Each example is followed by two English versions, the first a word-for-word literal gloss and the second a free translation of the meaning. For the Chinese examples in paragraphs, only a literal translation is provided. Terms and their abbreviations used in word-for-word literal glosses of Chinese examples are: classifier - CL; current relevant state le - CRS; modifier marker de - MM; negation word - NG; nominaliser le - NOM.

5. Themes are underlined in all examples.

6. The family, spring and autumn and The ordinary world are two Chinese novels.

7. http://www.cnnic.cn/en/index/index.htm

8. http://club.cat898.com/

9. Until July 2008.

10. http://bbs.sina.com.cn/

11. http://club.sohu.com/

12. http://tianya.cn

13. See Appendix A for a list of the six texts and their URLs.

14. http://club.cat898.com/newbbs/list.asp?boardid=26

15. The symbol // is used to separate clauses in this article.

16. Of the 243 interpersonal themes, only four are vocatives.

17. http://club.cat898.com/newbbs/dispbbs.asp?boardid=26&id=155669

Literature Review

Al-Sa’di, R. (2005). “Synchronous online chat” English: Computer-mediated communication. World Englishes, 24, 409-424.

Anis, J. (2007). Neography: Unconventional spelling in French SMS text messages. In B. Danet & S. C. Herring (Eds.), The multilingual Internet: Language, culture, and communication online (pp. 87-115). New York: Oxford University Press.

Biber, D. (1988). Variation across speech and writing. Cambridge, UK: Cambridge University Press.

Brown, P., & Fraser, C. (1979). Speech as a marker of situation. In K. R. Scherer & H. Giles (Eds.), Social markers in speech (pp. 33-62). Cambridge, UK: Cambridge University Press.

Chen, P. (1986). Referent introducing and tracking in Chinese narratives. Unpublished doctoral dissertation. University of California, Los Angeles.

Chung, D. S. (2008). Interactive features of online newspapers: Identifying patterns and predicting use of engaged readers. Journal of Computer-Mediated Communication, 13(3). Retrieved April 6, 2009 from http://www3.interscience.wiley.com/journal/119414160/abstract?CRETRY=1&SRETRY=0

Collot, M., & Belmore, N. (1996). Electronic language: A new variety of English. In S. C. Herring (Ed.), Computer-mediated communication: Linguistic, social, and cross-cultural perspectives (pp. 13-28). Amsterdam/Philadelphia: John Benjamins.

Crystal, D. (2001). Language and the Internet. Cambridge, UK: Cambridge University Press.

Danes, F. (1964). A three-level approach to syntax. Travaux Linguistiques de Prague, 1, 225-240.

Danet, B., & Herring, S. C. (2007). The multilingual Internet: Language, culture, and communication online. New York and Oxford: Oxford University Press.

de Oliveira, S. M. (2007). Breaking conversational norms on a Portuguese users’ network: Men as adjudicators of politeness? In B. Danet & S. C. Herring (Eds.), The multilingual Internet: Language, culture, and communication online (pp. 256-276). New York: Oxford University Press.

Duranti, A. (1985). Sociocultural dimensions of discourse. In T. Dijk (Ed.), Handbook of discourse analysis (Vol. 1) (pp. 193-230). New York: Academic Press.

Fang, Y. (1989). A discussion on the theme-rheme structure of Chinese. Journal of Tshinghua University (Philosophy and Social Sciences), 4(2).

Fang, Y. (2005). Systemic functional grammar and discourse analysis. Foreign Language Education, 26(6), 1-5.

Fang, Y., & McDonald, E. (2001). On functional structures in Chinese clause. Journal of Foreign Languages, 1, 42-46.

Fishman, J. (1972). The sociology of language. In P. P. Giglioli (Ed.), Language and social context (pp. 45-58). New York: Penguin Books.

Fries, P. H. (1995). A personal view of theme. In M. Ghadessy (Ed.), Thematic development in English texts (pp. 1-19). London: Pinter.

Gao, L. (2006). Language contact and convergence in computer-mediated communication. World Englishes, 25(2), 299-308.

Gao, L., & Yuan, R. (2005). Linguistic construction of modernity in computer-mediated mediated communication. In Q. Gao (Ed.), Proceedings of the Seventeenth North American Conference on Chinese Linguistics (NACCL-17) (pp. 66-83). Los Angeles, CA: GSIL Publications, University of Southern California.

Georgakopoulou, A. (2004). To tell or not to tell? Email stories between on- and offline interactions. Language@Internet, 1, article 1. Retrieved April 6, 2009 from http://www.languageatinternet.de/articles/2004/36/index_html/

Goutsos, D. (2005). The interaction of generic structure and interpersonal relations in two-party e-chat discourse. Language@Internet, 2, article 3.Retrieved April 6, 2009 from http://www.languageatinternet.de/articles/2005/188/index_html/

Halliday, M. A. K. (1978/2001). Language as social semiotic: The social interpretation of language and meaning. Beijing: Foreign Language Teaching and Research Press.

Halliday, M. A. K. (1994/2000). An introduction to functional grammar. Beijing: Foreign Language Teaching and Research Press.

Halliday, M.A.K., & Hasan, R. (1985). Language, context, and text: Aspects of language in a social-semiotic perspective. Oxford, UK: Oxford University Press.

Halliday, M.A.K., & Matthiessen, C. (2004). An introduction to functional grammar (3rd ed.). Oxford, UK: Oxford University Press.

Hasan, R. (1995). The concept of context in text. In P. Fries & M. A. K. Halliday (Eds.), Discourse in society: Systemic functional perspective (pp. 183-283). Norwood, NJ: Ablex.

Herring, S. C. (Ed.). (1996). Computer-mediated communication: Linguistic, social and cross-cultural perspectives. Amsterdam/Philadelphia: John Benjamins.

Herring, S. C. (2001). Computer-mediated discourse. In D. Tannen, D. Schiffrin, & H. Hamilton (Eds.), Handbook of discourse analysis (pp. 612-634). Oxford, UK: Blackwell.

Herring, S. C. (2007). A faceted classification scheme for computer-mediated discourse. Language@Internet, 4, article 1. Retrieved April 6, 2009 from http://www.languageatinternet.de/articles/2007/761

Hu, X-J. (1997). The thematic structure of modern Chinese. Unpublished Master’s thesis. Sichuan Foreign Studies University.

Hu, Z-L. (1994). The text cohesion and coherence. Shanghai: Shanghai Foreign Language Education Press.

Hu, Z-L., Zhu, Y-S., Zhang, D-L., & Li, Z-Z. (2005). An introduction to systemic functional grammar. Beijing: Peking University Press.

Huang, G-W. (2005). Theory and practice of discourse analysis: A study in advertising discourse. Shanghai: Shanghai Foreign Education Press.

Huang, Y. (1985). A preliminary discussion on the theme and rheme of English. Journal of Foreign Languages, 5, 32-36.

Hymes, D. H. (1974). Foundations in sociolinguistics. Philadelphia: University of Pennsylvania Press.

Jones, R. (2004). The problem of context in computer mediated communication. In P. Levine & R. Scollon (Eds.), Discourse and technology: Multimodal discourse analysis (pp. 20-33).Washington, D.C.: Georgetown University Press.

Jones, S. (1995). Understanding community in the information age. In S. Jones (Ed.), CyberSociety: Computer-mediated communication and community (pp. 10-35). London: Sage.

Jones, S. (Ed.). (1998). CyberSociety 2.0: Revisiting computer-mediated communication and community. London: Sage.

Lee, K. M. (2007). Linguistic features of email and ICQ instant messaging in Hong Kong. In B. Danet & S. C. Herring (Eds.), The multilingual Internet: Language, culture, and communication online (pp. 184-208). New York: Oxford University Press.

Li, N. (2006). A study on the thematic structure of modern Chinese. Unpublished Master’s thesis. Northeast Normal University.

Li, S-h. (2007) A systemic functional grammar of Chinese. London: Continuum

Liu, F-L. (2006). Thematic progression patterns in English to Chinese translation. Foreign Language Teaching and Research, 38(5), 309-312.

Martin, J. R. (1992/2004). English text: System and structure. Beijing: Peking University Press.

Murray, D. E. (1988). The context of oral and written language: A framework for mode and medium switching. Language and Society, 17, 351-373.

Nishimura, Y. (2003). Linguistic innovations and interactional features of casual online communication in Japanese. Journal of Computer-Mediated Communication, 9(1). Retrieved April 6, 2009 from http://jcmc.indiana.edu/vol9/issue1/nishimura.html

Nwogy, K., & Bloor, T. (1991). Thematic progression in professional and popular medical texts. In E. Ventola (Ed.), Functional and systemic linguistics: Approaches and uses (pp. 369-384). Berlin: Mouton de Gruyter.

Palfreyman, D., & Al Kbalil, M. (2007) “A funky language for teenzz to use”: Representing gulf Arabic in instant messaging. In B. Danet & S. C. Herring. (Eds.), The multilingual Internet: Language, culture, and communication online (pp. 43-63). New York: Oxford University Press.

Renkema, J. (2004). Introduction to discourse analysis. Amsterdam/Philadelphia: John Benjamins.

Su, H-Y. (2007). The multilingual multiorthographic Taiwan-based Internet. In B. Danet & S. C Herring (Eds.), The multilingual Internet: Language, culture and communication online (pp. 64-86). New York: Oxford University Press.

Thetela, P. (1997). Evaluation in academic research articles. Unpublished doctoral dissertation. University of Liverpool.

Thompson, G. (1999). Acting the part. In M. Ghadessy (Ed.), Text and context in functional linguistics (pp. 101-124). Amsterdam/Philadelphia: John Benjamins.

Wang, Y-F., Katz, A., & Chen, C-H. (2003). Thinking as saying: Shuo (‘say’) in Taiwan Mandarin conversation and BBS talk. Language Sciences, 25(5), 457-488.

Wu, Y. (2004). Spatial demonstratives in English and Chinese. Amsterdam/Philadelphia: John Benjamins.

Xie, B. (2008). Multimodal computer-mediated communication and social support among older Chinese Internet users. Journal of Computer-Mediated Communication, 13(3), 728-750. Retrieved April 6, 2009 from http://www3.interscience.wiley.com/journal/119414163/abstract

Xu, Y-C. (2004). A multivariate approach to context metafunction hook-up hypothesis. Unpublished doctoral dissertation. Fudan University.

Yang, C-H. (2007). Chinese Internet language: A sociolinguistic analysis of adaptations of the Chinese writing system. Language@Internet, 4. Retrieved April 6, 2009 from http://www.languageatinternet.de/articles/1142

Yates, S. (1996). Oral and written linguistic aspects of computer conferencing. In S. C. Herring (Ed.), Computer-mediated communication: Linguistic, social, and cross-cultural perspectives (pp. 29-46). Amsterdam/Philadelphia: John Benjamins.

Yates, S. (2000). Computer-mediated communication: The future of the letter? In D. Barton & N. Hall. (Eds.), Letter writing as a social practice (pp. 233-252). Amsterdam/Philadelphia: John Benjamins.

Zhu, Y-S. (2005). A dynamic study on context. Beijing: Peking University Press


List of the six threads, with URLs:

1. Where goes the future of Hong Kong movies《香港电影何去何从》 2006-3-28 http://club2.cat898.com/newbbs/dispbbs.asp?boardID=26&ID=1044559&page=1

2. Non-classical ‘Crazy Stone (a Chinese movie)’: why it is such a hit and highly complemented? 《“非经典”的〈疯狂的石头〉:凭什么疯狂,为什么赞美?》 2006-7-20 http://club2.cat898.com/newbbs/dispbbs.asp?boardID=26&ID=1196448&page=1

3. A comparison of the TV series of Japan and Korea 《关于日剧和韩剧的优劣比较(摘)》 2005-9-14 http://club2.cat898.com/newbbs/dispbbs.asp?boardID=26&ID=778234&page=1

4. ‘The Promise’ is just another failure: where is the future of Chinese blockbuster? 《〈无极〉此般〈英雄〉:“中国大片”何去何从?!》  2005-12-17 http://club.cat898.com/newbbs/dispbbs.asp?boardID=26&ID=902322&page=1

5. The tragic and subversive features of Leslie Cheung 《浅论张国荣的悲剧性与颠覆性》   2004-11-3 http://club.cat898.com/newbbs/dispbbs.asp?boardid=26&star=1&replyid=3982509&id=546816&skin=0&page=29

6. Pure love is almost impossible to a woman – a review to the movie ‘King Kong’  《女人很难拥有纯粹的爱情 —— 电影《金刚》之影评》  2006-1-20 http://club.cat898.com/newbbs/dispbbs.asp?boardid=26&star=1&replyid=9188487&id=953977&skin=0&page=39

Biographical Note

Xin Dai [daixinn@gmail.com] is an assistant lecturer in the School of Foreign Languages at the Southwest University of Political Science and Law in China. Her research interests include discourse analysis in computer-mediated communication, functional approaches to language, and language in popular culture.


Any party may pass on this Work by electronic means and make it available for download under the terms and conditions of the Digital Peer Publishing License. The text of the license may be accessed and retrieved at http://www.dipp.nrw.de/lizenzen/dppl/dppl/DPPL_v2_en_06-2004.html.