Abstract

Abstract A review of related works shows that ESP teachers generally perceive the abstract as a genre that is taught like the research article, and research focus has been on genre analyses in various disciplines. The important concern of ESP is to facilitate novice writers in producing abstracts for publication or attending conferences. A major question is whether abstracts used in the real world actually reflect prescribed conventions. On the other hand, information studies experts are mainly concerned with abstracts meeting the needs of end-users in information retrieval, and to this end whether abstracts fulfill standard measurements in information systems. A comparison of both perspectives appears to indicate the need for collaboration and awareness raising in both disciplines. ESP teachers need to examine the prospect of adding the complementary dimension of information science to the linguistic scaffolding that they have been providing to learnerwriters. A focus group discussion comprising professionals in the discipline of information science, whose expertise and interests include information retrieval, services, management, organization, and systems design, was conducted. The findings from the focus group corroborated that underlying the apparent differing viewpoints between both disciplines there are distinct interdisciplinary overlaps in the concept and objectives of the abstract and abstract writing. These conclusions have significant implications for future research and teaching practice. Introduction ESP teaching and research has been developing over the last thirty years. It began with a lexico-grammatical pedagogy for science and technical subjects, to genre issues in both the 80s and 90s, and arriving at the doorsteps of the new rhetoric Review of ESP perspectives In the 1960s, language teachers found themselves landed with the responsibility of instructing non-native English scientists and professionals to write acceptably for their new disciplines. Subsequently, these teacher-researchers, also known as ESP teachers-, based their pedagogy on the principle that language is for communication. The most popular practice was the genre approach developed and popularized by However, n ot as much interest has been placed on studying the abstract genre. Early writers who examined the abstract included 3 In the classroom, ESP teachers have used the abstract as a tool for their writing curriculums Recent research interest in the abstract reflected a more specific and deeper concern. The predominant questions include, How can non-native writers learn to write effective abstracts for their discourse communities Salager-Meyer (1990) studied abstracts from medical journals because of the frequent criticisms on badly written medical abstracts as being uninformative, misleading and lacking in internal structure. Results of her analyses revealed that many abstracts had no purpose statement and no conclusion and there was a prevalence of illogical sequencing in their move organization, and there were also flaws in the paragraph structuring and overlapping semantic concepts straddling between paragraphs. SalagerMeyer proposed that novice writers should be given good models to emulate. Santos (1996) studied how abstracts could be characterized in terms of their textual organization, and analyzed other key features of this genre. He selected 94 abstracts from three leading journals in Applied Linguistics. Santos found a prevalent five-move model with submoves: Situating the research, presenting the research, describing the method, summarizing the results, and discussing the results. He concluded that his resultant schematic pattern could provide pedagogical advantages to non-native learner-writers in reading and writing abstracts. Finally, Berkenkotter and Huckin interviewed seven scientists extensively to examine their research article reading habits. They found that these scientists practised a 4 scanning for interesting new i nformation strategy, for example they would read the title and then the abstract to 'size up' critical information and decide next on whether to read the full document. Next they selected a corpus of 350 journal articles to examine how research papers structure and information organisation have changed to accommodate present-day reading habits of scientists who have little time and who are faced with information overload. They found that titles and abstracts now tend to be more informative to cater to the n eeds or current readers. For example abstracts are now inherent parts of papers, and they have become longer and more informative, particularly in providing results and conclusions of studies so that readers may find it unnecessary to read the full document. From the studies above, it would appear that ESP interest in the abstract has increased in recent years and remains largely genre-related and understandably pedagogical. However, writers like Salager-Meyer and Posteguillo have overtly raised concern over the effects of technology on current speed of information flow, and Berkenkotter and Huckin have in their study observed how the abstract has become a significant surrogate document in information gathering. In view of this development among ESP experts it would make sense to corroborate their research with information science experts. Review of Information Science (IS) perspectives During the 60s and 70s, advanced technologies in information transfer methods have enabled a proliferation of information on the electronic information systems Consequently, information science writers paid great attention to producing guides to effective abstracting. They include writers who are experts on indexing and abstracting (Collision 1971; 5 Other information professionals followed up with institutionalized standards, the most significant being American National Standard Institute commonly called ANSI Z39. 14-1979, and International Standards ISO 214:1976. Such prolific literature on prescriptions for abstract writing belies the great concern of information professionals or experts in evaluating the quality of abstracts and on regulating the production of this surrogate A number of studies were conducted to assess readability using readability formulas, comprehensio n measures, or both. In more recent studies, information professionals (Wheatley and Armstrong 1997) have conducted several major studies on abstract production for online services, a direction to which the ESP researcher may need to shift. Finally, Pinto and Lancaster (1999) based their study on judging the quality of authored abstracts in terms of exhaustivity, accuracy, readability, cohesion, brevity, and cost. They found that although the computer has enabled easy availability of full texts in electronic forms this has not reduced the value of human produced abstracts. Motivation for Focus Group So far, the aforementioned review of related works seems to indicate that the content of abstracts is a major focus of information experts, and qualities like exhaustivity, accuracy, readability, organization and cohesion are deemed highly relevant. However, from the viewpoint of ESP, these qualities could speculatively be related to linguistic issues although at this point such conclusions are not readily apparent or distinct. In order to verify this speculation a focus group discussion was conducted among information science specialists or "specialist informants" after Bhatia and Dudley-Evans (Bhatia, 1993 and Dudley-Evans, 1986). The attempt to consult these specialist informants corroborates step number 7 of Bhatia's seven step process in analysing any academic or professional genres, a step which, apart from the works of Dudley-Evans and Bhatia, has not been practised more frequently as it should be in ESP. A total of seven information experts in the Information Studies Division at the School of Communication and Information Studies, 6 Nanyang Technological University participated in the study. The aim of the discussion was to uncover the actual nuances, opinions, recommendations and focus of the information experts with regards to the abstract as a tool in retrieval and dissemination. The Focus Group The participants are highly experienced professors who offer courses in library science, information organization and management, information retrieval and information systems design and programming at postgraduate level. Some of the participants are also journal reviewers and editors as well. Thus a comprehensive representation of information experts whose opinions, comments and recommendations would be highly reliable and respected is secured for this validation study. The objective of the discussion was to validate the claim that although there may be apparent differences in the way the language and information experts perceive, research, and qualify the abstract, there are in reality underlying similarities that can be better utilised to enhance abstract genre pedagogy for novice writers. The discussion was held in a specially designed and equipped room for focus group discussions. The room was equipped with both audio and video recording facilities, and a one-way glass partition that enabled unobtrusive observation during the live discussion. The discussion room accommodated a maximum of 12 participants sitting around a large square table (2mx2m) that had built-in microphones at intervals of about a meter around the table. Built-in cameras captured the images of the participants from all four angles. The discussion was scheduled for 90 minutes, and the entire session was video recorded. Three days before the session, a set of nine questions was given to the participants to outline the main issues of discussion (Appendix A). The thesis of the discussion was "Current technological advancements in information dissemination and retrieval have posed an additional challenge to Language providers. It is no longer enough to focus only on teaching rhetorical structures and linguistic conventions of the genre, instead, there is the need to meet the needs of writing abstracts for the WWW environment". Throughout the session, active participation among all members of the group was observed. Results and Discussion In the course of the discussion, some of the nine questions were found to be redundant, and unforeseen aspects of the abstract surfaced for discussion. The findings from the study are summarized and organized into six sub headings as follow: Setting the parameters for discussion: Defining the type of the abstract genre Although the first question on the effects of technology on the abstract was presented, the IS experts thought it was more appropriate to begin the discussion by setting some basic parameters. First it is important to consider the type and purpose of abstracts 7 under discussion. The type of abstract determines its product variations; there are descriptive, informative, critical, and review abstracts based on the different purposes and environments in which they are found. For example, the informative abstracts focus on results and conclusions and are preferred in technical documents; descriptive abstracts are shorter and merely provide abbreviated descriptions of the contents of the original documents, and critical abstracts go a step further to give an evaluation of the original documents. Similarly, it is as important to identify the purpose for the specific type of abstract selected. For example, an abstracting agency that aims to use the abstract to reach a global audience may decide that the descriptive type abstract is appropriate; on the other hand an abstract may be used in a book review and in this case the critical abstract is preferred. Whether the abstract under discussion is authored by the original writer, or with human assistance, or by machine extractio n it would be a pedagogical enhancement to create this awareness of typological distinctions in the classroom. Although this definition step is elementary, most ESP studies have so far focused mainly on the informative abstract. Another significant factor that could affect abstracts is the knowledge field. It is acknowledged by IS experts that different knowledge fields have different objectives for using abstracts. In this respect the abstract content and structures may vary. For example, abstracts in medical articles are structured under subheadings, a feature not practiced in other knowledge fields; abstracts in mathematical or chemical engineering fields use specific "unique identifiers" which are highly specialised terminology like mathematical symbols, but which are used to aid faster retrieval. What are the effects of technology on abstracts? Are electronic abstracts inherently different from traditional print-based abstracts? The IS participants are of the opinion that changes in environment are not responsible for affecting qualities of abstracts as long as the abstract purpose remains. For example, an authored abstract taken from a journal paper like IEEE Transactions or Association for Computing Machinery (ACM) and put into an abstract database, or even on an electronic platform like Internet does not need to be rewritten or adapted because it is the same abstract. Whether the abstract is used in print or Internet it should not impact or affect its value. However, at other times an original authored abstract could be modified by human assistance because of specific abstract or information agency objectives or policies. For example the abstract agency may want to modify original abstracts so that they reach a wider audience on the Internet to include the less technically trained readers. Electronic abstracts could be different because they have been generated automatically by machine, but these have no comparison with human produced abstracts and learners do not need to be concerned with them. Another reason that could impinge on the type of abstract is the result of overwhelming information flooding the systems and compelling the length and amounts of information in the abstract to be very brief and usually descriptive. 8 A point in future direction was brought up for discussion. In order to facilitate cross-referencing to full texts, online abstracts could have hyperlinks in various parts of the document to cater to different needs of the user. For example, if the user is interested in only the results of the study a link could be made to the result section of the full document. All these links are only a click away, whereas in the traditional environment, time and distance are constraints and such cross-referencing is not possible. However, the utility is possible provided the abstract in question is accompanied by its full document. [This also implies that the abstract content should cover all the various pertinent sections of the complete document for without such coverage, it would be impossible to create these hyperlinks to the affected sections] So far this idea has remained as a proposition. There was also the suggestion to include metadata such as author information, number of words, and publication details and other utilities as retrieval tools to assist the user in IR, but these have limitations and are thought best left to the professional indexers and other IR systems designers. What desirable qualities should abstracts have to increase their visibility and retrievability in IR systems? In an ideal situation, it is desirable for the abstract to be produced by a specialist in the knowledge field, who is also trained in abstract writing. However, this is very difficult to achieve and is very expensive. Instead, a more important consideration is to note the reasons why people use abstracts. Different users have different needs. This is especially so on the Internet where the abstracts reach a much wider audience, where some are experts in the field and others not. Based on this user consideration, readability and exhaustivity of content representation are important variables especially for researchers who use abstract as surrogates or as decision-making tools for reading source documents. The minimum quality is to aim for high readability in abstracts. Abstracts should be written in simple and direct language so that they are easily comprehensible to both experts and non-experts. In terms of exhaustivity, abstracts should contain sufficient and significant content representation to function effectively as surrogates. Finally, the practice of providing standard retrieval tools like keywords is also a means to enhance the abstract quality. It increases visibility and retrievability of abstracts on the retrieval systems. It would appear that the ESP or la nguage provider has the important dual-role of both providing the rhetorical structures and language conventions of the abstract and training novice writers to produce abstracts of high readability. Where exhaustivity and representation of content are concerned, engaging the collaboration of subject specialists could be a solution. Should abstract writers be taught the standard information retrieval tools used in IR systems? On whether abstract writers should be taught the standard retrieval tools used in IR systems, the IS experts were of the opinion that key words are important, although they 9 should not become the main objective for writing abstracts. For example, related terms like transputers and parallel processing should be represented in the text as a semantic group for greater audience reach through key word search. Semantic webs/thesaurus and tweaking of key words to fit retrieval needs could also be employed. On this point, a suggestion that links to these key words could be made to specific semantic groupings or semantic webs for users seeking on key word search alone was offered. However, the IS experts also pointed out one drawback in using all these features. Over enthusiastic writers may "abuse" the facility in order to ensure greater accessibility to their abstracts. Such activity would cause spamming problems on IR systems. One suggestion to counter this possibility was to ensure that abstract writing maintain stipulated standards in order to control or prevent such "abuse" by over zealous writers. The IS experts also stressed that only significant key words should be offered because it is only through using them that users can obtain correct or relevant information from the whole gamut of information available on the systems. Furthermore, in pioneering research and for new frontier knowledge, key words must first exist in the text before abstractors can make use of such a utility. To this end it is significant that the abstract is very important because it is here that new terms are introduced and communicated by authors, and later they may become the key words for search. On whether ESP teaching should incorporate indexing knowledge, the IS experts explained that indexing requires one to read the document technically, and to have the expertise to pull out the right information from the various parts of the document for the abstract. To do this professional indexers are required. Moreover indexers do not depend on abstracts alone. For example, indexers may have access to some unique identifiers that abstract writers may not have. Next, the use of rating utility is another facility available to assist the abstract user during retrieval. Reference to the rating of the abstract could assist the user in determining the relevance and significance of the full document. There was further suggestion that the rating utility is perhaps more important for abstracts on print than on Internet, one reason being users would be provided professional assessments of the quality of the abstracts, which would help them decide whether the full document is worth reading or purchasing. From the knowledge management viewpoint, having a standard rating tool is probably very useful, but it is very expensive and is not easily available. Moreover, even if one wants to provide a rating system the author cannot do it, because writers are very subjective in the assessment of their own writing. Such a system requires an independent professional body. This led to the general consensus that this utility is more useful in the filtering process later rather than at the process of writing. In summary, the implication is that abstract writers should have greater awareness of the search engine or the agency they are writing for. Writers should realize the agency objectives of using their abstracts, and the key words that are crucial and pertinent for 10 cross-referencing. However, the more technical capabilities like incorporating metadata, and rating utility should be left to the information experts. Thus, it would appear that the ESP practitioner's role is still linguistically based. How do IS experts rate the three aspects of an abstract: linguistic competency, content, and representation? Different viewpoints emerged in the discussion of linguistic competency. First, the term linguistic competency was clarified as referring to lexico-grammatical accuracy and use of language. To the IS experts it was clear that although they agreed that language is important it is the content representation that is more significant. This is especially true in the case of technical subject matter where the content is valued more than the eloquence of expression. To IS experts the communicative competence of writers is more important than merely language or grammatical accuracy. But to the editor, language is as important if not more important than the content in abstract. This is mainly because the abstract is a tool for communicating new information to users and as such the message must be succinctly and effectively communicated to achieve the purpose of its production. To this end writers must call upon their linguistic competence so that the content can work for them. Overall, the IS experts concluded that linguistic competency is important but should not take precedence over content representation of the abstract. The experts then identified several abstract-writing constraints: Brevity or word limit would affect exhaustivity of the abstract representation. Length and content structure would in turn depend on the type of abstract and agency requirements. Do main, journal, instructions by agencies would also limit an author's objectives of abstracting and the type of language depends on the technicality of subject matter. While these constraints are outside the jurisdiction of the author, readability and linguistic qualities, which are important aspects of the abstract, are within the author's control. Simplicity is the key to writing readable abstracts. In order to improve the readability the writer would need linguistic skills. Linguistic skills would impinge clarity, conciseness, and precision, while ensuring good cohesive and coherent text relationships at the same time. A point was raised that often writers do not present content exhaustively and the reason cited was lack of linguistic competence. There is also the need not only to know what input information should be included in the abstract, but that it be presented communicatively and meaningfully to the user. Thus, writers must consider the uses and needs of readers. One of the experts went as far as to say that an abstract must be as good as if not better than the full document because of its surrogate significance. When asked to rate their opinion on which aspect of the abstract they would consider very important on a scale of 1-3 the consensus was returned in the following order beginning with the most important. 1. Content: By this rating they emphasized the significance of content presentation and content representation of information in the full document.