Discourse organisation in computing science discourse: a preliminary study

Discourse Organisation in Computing Science Discourse: A Preliminary Study Francisco Alonso Altneida-Laura Cruz GarcíaVíctor González Ruiz Universidad de Las Palmas de Gran Canaria The present work aims at describing the structure of computing science texts in a very limited time-span. These texts show a similar organisation of information; and we have also noticed the presence of certain linguistic elements which we consider generic of these texts. We have also concentrated on the visual elements which accompany the main text. These visual elements are classified according to their form. 1. The aim of study This paper aims at describing the way how information is conveniently organised in computing science popular papers, such as those contained in computer magazines like Communications of the ACM, and Dr. Dobb's Journal. Our study will focus on highlighting the schematic structure of those papers reveaUng the grammar frequently associated with those stages in the construction of computing science discourse. A detailed description of the visual aids used to guide the reader through the text will be also offered. 2. Material of study The material used for this study consists of fourteen articles published in the latest issues of different computer science magazines: Communications of the ACM, PC Magazine, Dr. Dobb's Journal, Byte and C++ Report. They have an extensión which ranges from about 1500 to 4500 words, in order to show that the organisation of information does not necessarily depend on the extensión of the text. The articles normally deal with a certain computer product, its advantages and disadvantages, its evolution and the new improvements, and also and quite frequently the similarities and differences between the given product and other products of the same kind. Revista de Lenguas para Fines Específicos N""* 5 y 6 (1998-1999) 15 Francisco Alonso Almeida-Laum Cruz García-Víctor González Ruiz We have decided to work on the latest issues (from November 1998 until January 1999) of the corresponding magazines. The reason for this choice lies in the fact that we are not very much interested in the evolution of the informative structure of this type of discourse, but in finding features which are common to a certain number of articles belonging to the same period of time, in order to describe the way information is structured and presented and, thus, set a general pattern. The journals selected are directed toward a specific audience formed by computer science experts and users with a certain level of knowledge of the matter. They are not research papers, since they give general information about a computer product and do not intend to show any research academic study. The authors of the texts are usually editors, professors, programmers and researchers. 3. What's a popular computing science text? Popular computing science texts basically aims at presenting new information about a specific product, showing how that specific product works, and very often pointing out the advantages of that product with respect to a former one. Obviously, this sort of information is addressed to a specific sort of audience. This audience constitutes a parameter to take into account when defining the concept of popular computing science text. Also, the formal characteristics of these texts suggest the writer's preoccupation to achieve clear messages which the reader may comprehend effortless. Within the formal characteristics, it should be noted the presence of certain typographical elements which guide the writer in his search to find certain pieces of information. The audience of these texts is basically constituted by users, and people with a certain degree of knowledge. The consequences of this idea is that certain pieces of information should necessarily be ignored, since they may prove redundant for these types of readers that own a wide knowledge of the topic under discussion. 16 Revista de Lenguas para Fines Específicos N"" 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminanj Study Obviously, this does not imply that the writer includes a background introduction which shows the antecedents to the product which he is about to describing. What it does really mean is that the writer deems unnecessary to describe former products in detail. Would the reader need that sort of omitted information, he should direct himself to books oriented to the explanation of those particular products. Those books normally accompany the purchased product to describe it, or they might be bought at any bookshop. Apart from conditioning the amount of information to be included in those texts, the audience also conditions the type of language used there. In other words, the writer makes efforts to describe the products using clear language departing from excessive gilding strings, and complicated subordination which may obscure the information the article deals with. This is specially relevant if we consider that many of these texts describe computing programmes which, to some extend, also require the use of mathematical elements which may posit some complications by themselves. So, why drawing the readers' attention to untangle difficult language when the aim of the paper is to illustrate the pro's and, rarely the con's, of new products? A more formal definition of these texts involves the use of certain grammatical features which appear recursively in the different texts under study. These features which we label as generic of this sort of texts are (1) the use of the present simple tense, (2) expressions of reason, purpose and result, (3) the presence of causative verbs, (4) the use of comparative and superlative adjectives and adverbs, (5) the massive presence of can, and (6) the construction by + -ing. It could be argued the convenience of including in the above list the presence of technical vocabulary in those texts as a way to recognise them. However, we think the specific type of vocabulary used here is also part of other computing science papers, and thus it does not prove generic of these texts. Revista de Lenguas para Fines Específicos N""* 5 y 6 (1998 y 1999) 17 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz What follows is a description of the visual devices used in order to guide the reader through the text. Secondly, we offer the schematic structure of the texts studied here, and finally, the recursive linguistic elements in these texts are presented. 3.1. Visual devices in computing science popular papers Visual devices are normally used in this type of texts as a way to highlight specific bits of Information, some times. Other times, these devices constitute mere summaries describing some aspects related to the product presented. The nature of these visual devices are varied, but we can summarise those in (1) typographical elements such as bold type-face script, the use of bigger fonts, dropped capital letters, and the like; (2) in-text Information boxes such as schemas, drawings, flowcharts, and so on. 3.1.1. Typographical elements Typographical elements are used in our texts to signal how Information is divided into different stages so that the reader can find in a reduced amount of space the Information they require. For instance, the author of the text Get the Message systematically employs the bold type face in order to signal the beginning of the description of a new product: Telelogos Media Transfer Media transfer, from the French firm Telelogos, .... (p. 8) XcelleNet RemoteWare With over 650,000 users in 1700 organizations, Xcellenet is the leading vender... (p. 9) Mobile Objects Mobile Objects differ from Media Transfer and ... (p. 9) The Wireless Future Mobile Computer users will increasingly turn to wireless telecommunications... (p. 11) 18 Revista de Lenguas para Fines Específicos N""^ 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminar}/ Study In the case of other authors, they combine the use of bold-type face with italics as a way to indícate subdivisions of the information, as can be seen in the following example from The Virtual Design Team: The VDT Micro Theory of Project Organizations Organizational Engineering is the process of configuring an organization structure... (p. 85) VDT Implementation The VDT model uses... Acíivities. VDT shifts the object of analysis from an aggregated organization... Communications. Coordination requires information flow... Actors and information processing. Because of its aggregated view of organizational information... Allocate attention. Activity subactivities and Communications accumulate... Process information. After an actor selects an activity or coordination Ítem... Send Communications to other actors. Actors use communication... Genérate and handle exceptions. VDT actors genérate, communicate, and process several... Exceptions and decisión making. Actor information processing and exception... Programming implementation. The VDT system was implemented as an object-oriented... Discussion As discussed in the procurement policy case study ealier, we used VDT to .... (pp. 85-91) In the above example, the author uses the bold-type face and italics as a way to signal the división of his text. In doing so, his achievement is twofold: (1) the división of the text into different units of information, and (2) the presentation of his text as readerfriendly. The way he uses those two typographical elements is by Revista de Lenguas para Fines Específicos N°" 5 y 6 (1998 y 1999) 19 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz highlighting sections in bold, and subsection in a combination of bold type-face and italics. Further subsections included in subsections are indicated by the use of italics as shown above. This technique also allows the writer to avoid excessive wording by tackling directly those aspects whicli must be necessarily discussed. This is specially relevant if we consider the small amount of space devoted to each article in each magazine. In other words, articles are only few pages length, what implies that the writer must try to condense all the information he wants to pass on to his readers in clear stages. 3.2.2. In-text information boxes We understand by the concept of in-text information boxes to those schemas, figures, drawings, and flowcharts which are included in our articles as a way to clarify certain important points dealt with within the main body text of the articles. Sometimes the in-text information boxes are used to include the stages of conclusión and further information. In-text information boxes constitute an undeniable help to the reader, mainly because they can resort to them when they find the message obscure or when they need a drawing to follow the author's explanation. An example of schema is the following {Get the Message, p. 8): 20 Revista de Lenguas para Fines Específicos N°^ 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminar]/ Study Messaging Middleware Application Presentation Messaging Middleware Session Transpon Network Data Link Builds and controls connection. Rebuilds connection if line drops. Recognizes which data has been transmitted and which has not; restarts transmission exactly at the point of failure. Replicates databases Ensures security, access and rights. In the example above, the writer outlines the qualities of the product in a sort of schema. By doing so, he facilitates the reader's comprehension of the message, partly because the language he uses at this point is far more simple than in the body text. In addition, the presentation of the general features of this product in such a clear outline allows the reader to check in seconds if the product would be of any benefit to him, so that readers can eventually save their time. Another example of in-text information boxes is found in the same article Get the message (p. 8). Here, the author presents a checklist comparing the features of the producís he describes in his article. Tables like this help the reader to have a complete view of the products described in the article in a very limitad period of time.: Revista de Lenguas para Fines Específicos N"'' 5 y 6 (1998 y 1999) 21 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz Product Store and forward Error recovery Data compression Syncronization Custom coll. hand. Subscr. management Asset management Mixed transports V = yes Middleware Features Checklist Media Transfer V V V V V V RemoteWare V V V V V V V Mobile Objects V V V V V Nokia WDE V V 4. The structure of texts 4.1. División of texts in sections Most computing science popular papers contain an almost identical organizational pattern. The reader can find the same textual stages in nearly all the articles under study. However, there exist several varying factors, such as the section of the magazine where the article is in (which determines where the focus must be on), or the author of the text (whose individual style may modify the relatively fixed structure). A generic pattern of organization for this type of texts is outlined below: 22 Revista de Lenguas para Fines Específicos N"" 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminary Study the author is doing (or, at least, intends to do) in the following pages. But the reader does not always avail of these markers expressing the aim of the article. Sometimes what the writer intends to achieve is expressed through a rhetorical question normally contained in the first paragraphs of the body text. It must be understood as a question which will be answered throughout the article, like in the following example, in which the lack of an immediate proper answer indicates the reader that the question is rhetorical and that it relates directly to the purpose of the paper. How does so small a package accomplish so much? {Palm Computing: So Much froin So Little, p. 263) Sometimes, though, there is not even a rhetorical question that can help the reader find out the purpose of the paper. When such markers do not exist, the reader will still be able to combine loóse pieces of Information throughout the introductory section in order to make up what the writer intends to do in the article. Most times this Information can be also inferred from the contents of other sections, such as the title, the preand post-title headlines, the contextualization, and even the ñame of the author and the section of the magazine. The article C++ Containers Are Not Their Elements, for instance, does not explicitly expresses the purpose of the paper in the introduction. Notwithstanding, the combination of different sentences from the contextualization, as well as the ñame of the section where the article is inserted {Traps and Pitfalls) help the reader determine this Information. The contextualizing parts referred to above are: In a previous article 1 [this superscript number makes reference to a final note where details of this previous article are given], I pointed out a commom mistake that programmers make while learning how to use the C++ standard library. {C++ Containers Are Not Their Elements, p. 13) Revista de Lenguas para Fines Específicos N°* 5 y 6 (1998 y 1999) 29 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz The naive use of copy fails because iterators make available only a limited subset of the possible container operations. These limitations constrain not only uses of iterators, but also uses of algorithms. (C++ Containers Are Not Their Elements, p. 13) Bearing in mind that these are the initial and final sentences of the introductory section of this article, usual readers of this kind of magazines will not take long to infer that the purpose of the paper is to give suggestions in order to avoid some errors related to the programming of the C++ standard library. 4.1.4.1.2. Contextualization The contextualization, although a priori an integral part of the introductory section, may sometimes appear inserted in several other sections of the paper, since the different stages in the description of a product, for example, may have their own contextualizing sub-sections. However, here we are referring only to the general contextualization of the topic to be dealt with in the article (a simple description of a product, a comparison of two products or a problemsolution type of text'). The contextualization section normally offers brief background Information about the development of the product to be reviewed in the article, or some comment on the current situation of a computerrelated topic and how it has developed since it was first considered. The frequent use of time and place expressions (in order to physically sitúate what is being described) marks this section, as these examples show: The 3Com Palm III personal digital assistant (PDA) and its older siblings have become increasingly popular since the system was introduced thanks to their simplicity, functionality, and ease of use. {Palm Computing: So Much from So Little, p. 263) For a detailed description of this type of texts, see Hoey (1994:26-45). 30 Revista de Lenguas para Fines Específicos N°'' 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Disconrse: A Preliminar]/ Study As with all computing machinery, the Palm relies on an operating system for its functionality. Like a networked PC, which includes both an OS and a means of linking to the network operating system, the Palm includes both the OS (currently Palm OS, Versión 3.0) and software that lets the device link with a PC or a Mac. {Palm Computing: So Much from So Little, p. 263) What better place to do battle with your rivals than on their turf? This is the approach that Compaq Computer Corp. is taking with the mid-November introduction of its Prosignia line of desktops, notebooks, and servers for small businesses. {Compaq Takes Aim at Direct Vendors, p. 56) In the first example, the author briefly introduces the product which is going to be reviewed and gives account of its successful development. In the following one, the product referred to in the title is presented and defined with respect to other computing devices, so it can be also considered a way of contextualizing. Finally, in the third example, the date of the introduction of the new Prosignia line marks the beginning of a fictional «battle», which will be represented in the article by means of a comparison of these new Prosignia products and their major competitors. 4.1.4.2. Product description Although most computing science popular papers follow a same generic pattern or schematic structure, not all of them are devoted to the same type of Information. Some of them merely deal with the presentation of the features of a new product; some compare two rival products (or an upgraded product with its older versión); and others present a problem-solution text type, where a more or less general computing problem is introduced in the first place, and then a suggestion is given in order to solve this problem. We have used the generic phrase «Product description» for this section as all of the anieles studied, although slightly differently organized according to which of the three types of Information they contain, have the analysis of one or more products as a defining feature. Revista de Lenguas para Fines Específicos N™ 5 y 6 (1998 y 1999) 31 Francisco Alonso Alineida-Laura Cruz García-Víctor González Ruiz Now let US observe how the different types of information determine the way the text is organized. It must be taken into account that authors very frequently combine the three types of information, thus producing mixed texts and putting together some of the characteristics explained below. Presentation of the features of a product Although it is very unlikely to find a text where the features of a product are presented and no reference is made to other products or to generic computer-related problems, there exist some articles where the focus of the paper is mainly on the detailed description of some particular machine or device. These texts are normally divided into sub-sections (preceded by headings which determine what information is given next). For instance, in the article Palm Computing: So Much from So Little, details about memory and data storage, power modes, graffiti and hotsync and conduits (all features of the new Palm personal digital assistant) are presented consecutively, providing the reader with a general picture of the machine. Some excerpts from this article show the recursive use of typical linguistic features: The device's ROM holds the operating system plus the applications that ship with the unit; RAM is used for any applications you install or develop yourself, user data for all applications, and any other user-configurable data. Because the RAM is powered by the Palm's batteries, data is safe when the Palm enters low-power modes (see «Power Modes» below). You will lose data in RAM if the Palm is without power for over a minute, however, so it's important to transfer your data to your PC before losing battery power (as when you change batteries). (p. 263) Switching from one application to another is instantaneous, because primary memory simply closes the first and launches the second, with the modified data also stored in primary memory. In other words, data never leaves the nonvolatile primary memory in which it is stored, so there's no need for an external storage device in which to preserve it. (p. 264) 32 Revista de Lenguas para Fines Específicos N'"= 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminary Study Present simple tense verbs, conditional constructions (expressing a cause-effect relation), series of alternative possibilities (linked by means of commas and conjunctions), and expressions intended to facilítate and clarify the reading {so it's important to, in other words) are some of the linguistic features employed by the author in order to present the product reviewed as clearly as possible. Compañson of two producís Some sections in computing science popular magazines are exclusively devoted to the description of new products (e.g. First Looks in PC Magazine). The authors of the articles included in these sections normally compare the new products with other releases of the same type, or with older versions of the same machines. The following example presents the linguistic features used recursively by these writers: Comparing one vendor's non-corporate-focused desktops with another's —strictly on the basis of hardware— is a relatively easy task, as most use the same core components. And that's what we found when comparing the Prosignia Desktop 330 with a Dell Dimensión XPS R450. {Compaq Takes Aiin at Direct Vendors, p. 56) Subtle differences do exist. Whereas the Compaq PC uses a 16.8GB hard disk, the Dell can be ordered with a new 17.2GB disk. Both have speakers in their standard configurations, but the Dimensión's speakers (an Altec lansing ACS295 system with subwoofer) have superior sound. You can, however, order the Prosignia with Altec Lansing ACS48 speakers and subwoofer for an additional $85. Both machines are mini-towers, but the Prosignia can be reconfigured as a desktop; the Dimensión, on the other hand, provides easier internal access via a single thumbscrew. {Compaq Takes Aim at Direct Vendors, p. 56) The use of comparative adjectives {superior, easier) and expressions {comparing ... on the basis of, when comparing), nouns denoting «difference» {differences), and the «alternating» style of the text (similar features of the two products reviewed are alternately descriRevista de Lenguas para Fines Específicos N" 5 y 6 (1998 y 1999) 33 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz bed with the help of linguistic markers, such as whereas, however, on the other hand) help the reader notice that the author is comparing two different products. Problem-solution text type Computer-related problems are constantly referred to in the articles under study. Most times they appear in order to highlight the advantages of a product (as long as certain features of the product can handle those problems), and are not given an in-depth treatment. On some occasions, however, the body text is organized as a twostage structure, where the first stage is the presentation of a more or less general problem, and the second one, the proposal of available Solutions. In the article Get the Message, for instance, approximately a quarter of the body text is devoted to explaining how much the difficulties of remote communication through networking software can affect the everyday transactions of all kinds of a company. An example of this problem stage can be observed in the following excerpts: Dial-up Communications is still something of a black art, involving start-up scripts, Winsock settings, and other unfriendly stuff that can make Ufe hell for a salesman in a hurry in a motel room. (p. 7) «The Internet infrastructure is far from sufficient for supporting business-critical transactions,» says Rosemary Rock-Evans, an associate at the U.K. consultancy Ovum. «Web browser/Web server software simply does not provide the scalability, availability, reliability, security, or high performance that such Services need.» (p. 7) The weakness of such solutions usually is the software. Any solution that relies on ordinary Windows Communications products is likely to be too difficult and too unreliable for most salespeople. Then there's the problem of navigation once a user is logged on, finding the right applications and databases and issuing the appropriate queries. (pp. 7-8) Certain linguistic features, such as the use of negative-meaning words (adjectives: black, unfriendly, difficult, unreliable; nouns: hell, 34 Revista de Lenguas para Fines Específicos N°' 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminary Study weakness, problem), phrases (far from sufficient) and sentences {Web browser/Web server software simply does not provide the scalability, availability, reliability, security, or high performance that such services need), as well as the style, focused mainly on the practical consequences of the problem, are to be considered as defining characteristics of this stage. After the problem has been explained, one or more available solutions are presented. In the article used as example {Get the Message), the author offers several products (divided by headings) and evaluates the way each of these tackle the Communications problem (the characteristics of this part of the body text are as shown in Presentation of the features of a product above). However, hints of a solution are already included throughout the problem stage, as can be observed in the following excerpts: That's why a new class of message-oriented middieware is appearing, designed precisely to iron out the uncertainties of remote networking. (p.7) This approach is particularly helpful when you're using the World Wide Web. (p. 7) The result is much quicker queries, which can be very important when you're sitting face-to-face with a customer. (p. 8) The great variety of positive-meaning linguistic elements (words, phrases and whole sentences) and the highlighting of advantages of a new group of products, contrast with the negative context (as explained above) where these excerpts are inserted. 4.1.4.3. Conclusion/Evaluation Authors of this type of articles usually give a final evaluation of the products they have described in the text. This sort of conclusión, which does not always appear in such a manner, normally emphazises the advantages of what has been reviewed, including the autRevista de Lenguas para Fines Específicos N'" 5 y 6 (1998 y 1999) 35 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz hor's own valué judgments and indirectly appealing the reader to buy the product, as the foUowing examples show: As more GSM operators introduce services like AWDE, the cellular network will become a more useful means of data Communications. Liberating the sales forcé from office and hotel telephones is useful in the industrialized world. In countries where the public telephone infrastructure is sparse or nonexistent, it's a necessity. (Get the Message, p. 11) As you can see, the Palm OS ensures that this little PDA manages data properly. It allows fast, easy, and usable input, and it synchronizes efficiently with desktop software running on the PC. No wonder the device has won the hearts and minds of so many. {Palm Computing: So Much from So Little, p. 268) This —in some cases exaggerating— final account of advantages may, however, not appear at the end of the article (as in the examples above). Sometimes it could even seem that there is no conclusión at all, since no linguistic markers (such as lexical items denoting summarizing or global evaluation, phrases like as you can see, or the change from a merely informative style to a more opinionoriented one) are to be found in the papers. Readers can then resort to the title or pre— and post-title headlines (which could include some kind of evaluation); some figures' captions (where a summary of the article is frequently offered); or the in-text Information boxes (which are often used by authors to give their own views far from the technical data of the body text). 4.1.4.4. Reference to further information Unlike research academic papers, computing science popular articles rarely include bibliographical information. The type of referenees that authors of these texts most frequently provide the reader with is practical details, such as telephone numbers and Web site addresses related to the product or products reviewed and the companies that produce them. This information, usually contained in 36 Revista de Lenguas para Fines Específicos N" 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminary Study marginal notes and in in-text information boxes (not necessarily divided from the rest of the body text), can be presented in a schematic way, as shown in the following examples: Compaq Prosignia Desktop 330. $2,239 direct. Compaq Computer Corp., Houston; 800-888-9909; www.compaq.com/smb. (Compaq Takes Aim at Direct Vendors, p. 56) Ñame of product; price; ñame of company, its telephone number and its Web address, are the practical data offered to the readers interested in further contact and information. The following linear diagram represents the schematic structure which we have described stage by stage above (under Eggins 1994: 40): (Ñame of section of magazine) ^ (Pre-title headline) ^ Title '^ (Posttitle headline) ^ Ñame of author/s * Purpose of paper * Contextualization ^ Product description '^ (Conclusión) * (Reference to further information) Key: '^ indicates fixed order, * unordered stage, ( ) optional stage 5. Linguistic features in texts After the description of the structure of the texts, we will pay attention to certain linguistic features which co-occurred in all of them. Those features are the use of the present simple tense, comparatives and superlatives, expressions of reason, purpose and result, causative verbs like allow, let and enable, the modal verb can and the construction by + -ing. 5.1. Use of the present simple tense As in this sort of texts there is a description of a product, it is usual to find a great deal of verbs in the present simple. This tense is used for explanations and descriptions on how something works. Revista de Lenguas para Fines Específicos N°" 5 y 6 (1998 y 1999) 37 Francisco Alonso Almeida-Laura Cruz García-Víctor González Ruiz what it can do and what it is for. It frequently appears in explanations with expressions like for example and for instance: The remote run command works via «robot» programs that can control existing applications by feeding them keystrokes... {Get the Message, p. 8) When the stylus touches the Graffiti display área, the event manager calis the pen manager and informs it of the pixel location. {Palm Computíng: So Much From So Little, p. 268). Again, as soon as the Palm detects user input it comes out of sleep mode and goes directly into running mode. {Palm Computing: So Much From So Little, p. 268). The use of the present simple tense in the first example {works) obviously aims at explaining how something {the remote run command) works; and, in the second and third examples, at describing the process of what happens when a certain action is carried out or something takes place {touches, calis, informs, detects, comes, goes). The actions are expressed by means of this verb tense since they always occur. 5.2. Expressions of reason, purpose and result Due to the informative nature of the sort of texts we are dealing with and the contents they present, the relations of reason (cause and effect, reason and consequence), as well as expressions of purpose and result appear very frequently. 5.2.1. Reason relationships Cause and effect Cause may be indicated by an adverbial ¿ecaiííe-clause, by a prepositional phrase using the expressions because of, from, out of, by some verbal constructions like lead to, result in, give rise to or by some causative verbs. However there are other ways of expressing this relation as we will show in the third example below: 38 Revista de Lenguas para Fines Específicos N"" 5 y 6 (1998 y 1999) Discourse Organisation in Computing Science Discourse: A Preliminary Study The above study stands as a first step towards the analysis of computing science magazine articles as a specific text-type, and henee, there are many aspects which have not yet been covered within the framework of this paper. Future research in this área must include information at a more pragmatic level, such as the use of language intending direct addressing to the reader, or even the presence of what we have called here in-text information boxes, which we felt should be studied on their own. WORKS CITED Primary Sources Freed, L. 1999. «Boot Up, Turn On, Tune In» in PC Magazine, January 5: 5657. Hummel, R. L. and D. Andrews. 1998. «Notebook Screens to Get Bigger and Better» in Byte, July: 20-21. Kao, S. 1998. «Your PC Never Sounded So Good» in Byte, July: 13-19. Koenig, A. 1998. «C-I-+ Containers Are Not Their Elements» in C++ Repon, November-December: 13-14. Kunz, J. C. et al. 1998. «The Virtual Design Team» in Communications of the ACM, Vol. 41, No. 11: 84-91. Martin, R. C. 1998. «UML Use-Case Diagrams» in C++ Report, NovemberDecember: 49-51. Moore, R. et al. 1998. «Data-lntensive Computing and Digital Librarles» in Communications of the ACM, Vol. 41, No. 11: 56-62. Pountain, D. 1998. «Get the Message» in Byte, July: 7-11. Randall, N. 1999. «Internet Apps in Win 98/NT 4» in PC Magazine, January 5: 241-242. Randall, N. 1999. «Palm Computing: So Much from So Little» in PC Magazine, January 5: 263-268. Shankel, J. 1999. «Little Languages with Lex, Yace, and MFC» in Dr. Dobb's Journal, January: 28-33. Tabibian, O. R. 1999. «Sun Ups the Ante» in PC Magazine, January 19: 75. Venezia, C. 1999. «Compaq Takes Aim at Direct Vendors» in PC Magazine, January 5: 56-7. Webb, P. and G. V. Wilson. 1999. «Matlab as a Scripting Language» in Dr. Dobb's Journal, January: 94-99. Revista de Lenguas para Fines Específicos N°^ 5 y 6 (1998 y 1999) 45 Francisco Alonso Almeida-Laiira Cruz García-Víctor González Ruiz Secondary Sources Eggins, S. 1994. An Introduction to Systemic Functional Linguístics. London: Pinter. Hoey, M. 1994. «Signalling in Discourse: a Functional Analysis of a Common Discourse Pattern in Written and Spoken English,» in Coulthard, M. (ed) Advances in Written Text Analysis. London: Routledge, 26-45. Leech, G. and J. Svartvik. 1975. A Conimunicative Grcimmar of English. Longman. Quirk, R. et al. 1985. A coniprehensive Granirnar of the English Language. Longman. 46 Revista de Lenguas para Fines Específicos N™ 5 y 6 (1998 y 1999)