No abstract

No abstract

There are two major problems in the field of information use today—the quantity of material produced and its quality. Quality is of two kinds—innovation or newness of the results, and the validity of the results. This paper is concerned with the problem of validity and how to screen the literature using this criterion as a filter. Various quality filtering systems are discussed, including peer review and expert panels, and the problems associated with them delineated. It is proposed that the information analyst (defined anew herein) be used as a filter to identify quality research papers, especially using the validity criterion. A discussion of the pros and cons of the recommendation is initiated. The paper has two objectives: (1) to establish a definition of one of the jobs in the information field, and (2) to present an idea that will increase our ability to identify and control the quality scientific literature. The growth of literature and the question of qualityGoffman and Warren's [13] Scientific Injormation Systems and the Principle of Selectivity addresses two issues in the study of the nature of knowledge that are rarely if ever touched on by historians and philosophers of science. These two issues are peculiarly in the field of study of librarianship and information science; of librarianship because of the implications for practice, and of information science because it is information scientists who are responsible for discovering, testing, and describing solutions to problematic phenomena in the use of information and knowledge. The issues are: (1) the growth patterns of the knowledge record, that is, of the literature, and (2) the quality of the collection, or the library, gathered from that record for use.To study the growth of the record, information scientists have developed the apparatus of biblio-metrics. To ensure the quality of collections, librarians have identified general criteria such as those delineated in Sheehy [31] and Higgins [16].Our concerns about the record have two bases.One is the responsibility that librarianship has always felt, as an ideal, to collect and store the complete output of recorded thought regardless of quality (because future usefulness and values are unknown), in order to preserve the cultural heritage. Of course, that ideal is impossible to reach. The problem is twofold: (1) storage space, and (2) the discovery and control of all publications. The second concern is to select from the known record, on some criteria of quality, for current use. The problem here is not only deciding on the criteria, but also applying them.The information we need regarding the growth of the literature is the number of items that presently exist and that we will be facing in the future, and what patterns, if any, growth follows. The information we need regarding quality is how to make judgments about quality that will enhance the success of a user's personal programme. This

Farradane's categories of relations (Fig. 1) are viewed as percepts rather than concepts. It is argued that Farradane's original use of language supports this view. A comparison of Farradane's categories with perceptual discriminations in humans is attempted. The conclusion seems to support claims made for relational operators, whether those of Farradane or similar relational indexing devices as in PRECIS, to have the potential to act as metalanguages.

Evidence is presented to show that the body plays a continuing and fundamental part in the process of thinking and the development of the knowledge used to think. The conclusion is that, if this is true, then there is a crucial part of our subjective knowledge base which cannot be represented in an objectified scheme, whether for artificial intelligence (AI) or information retrieval (IR), and therefore we cannot expect to get a computer to think as a human being thinks. THINKING SOME YEARS AGO I published a brief description of ways in which 'scholars' thought which did not involve the use of words [1]. One example showed how a historian could reconstruct an event in his imagination and, as an actor in the scene, go through the thinking processes of the historical figures. Another described the work of a microbiologist as he rotated the image of a cell in his mind and, as a spectator, 'saw' the missing parts. A third example was a little different as it told how a wide receiver in football studied slow motion movies of an opposing team to get a sense or feeling of where the defence would be at different times so he could react without thinking.At the time, my conclusion was that the use of conceptual models of knowledge (indexing systems, trees, networks, frames, taxonomies) to represent knowledge in IR systems could be restrictive and certainly incomplete because they were missing too much of the information in