back next next
Table of Contents
Index
Glossary
































































































Table of Contents
Index
Glossary
back next next

Next: Representing scientific discourse relations Up: A discussion of the Previous: A discussion of the

  
5.4.1 Representing organisational relations

In order to prevent the reader from getting lost in the network of modules that is formed by a modular article or a larger collection of modules, the organisational coherence has to be made explicit.

As an indication, we list in table 5.4 how many relations of the different types we have identified for each content module in the modular version of A08, and how many links we have created to express these relations.5.20 In A08, we identified, and expressed in explicitly labelled links, 377  organisational relations. The links provide the reader with roads to the various modules, the labels serve as signposts, and the module Map of contents as a road map.

The identification of an organisational relation usually did not lead to the creation of an explicit link in the text of the content modules.5.21 However, when we created a link to connect objects that were related by a scientific discourse relation, we also expressed the organisational relations between those objects in that same link. Thus, the reader is informed of the relation between connected objects, so that he can make a well-considered choice as to whether to follow the link or not. In the modular version of A08, organisational relations make up 40% of the complete set of relations that are expressed in the links: 18% are included in the text and 22% are only made explicit in the navigation menu.

 Creating links expressing hierarchical relations, we aimed to give them access to the constituent modules of a complex module, and vice versa, to the covering complex module. Hierarchical relations automatically appear when  complex modules are formed from constituent modules. The only links that can express a hierarchical relation, and thereby the fact that the complex module contains the constituent modules, are ones which connect modules that are also close according to the proximity-based relations. If one of the modules is microscopic, and thereby part of an article, they both have to be part of that article, so that the link is also characterised as `internal'.  Mesoscopic and macroscopic  modules that are connected by means of a link carrying a hierarchical label, are also part of the same (higher-level) entity.

In the examples, all  elementary modules are part of a larger  complex module, and in all complex modules the constituents are summarised (even if only in terms of a table of contents). Therefore we identified hierarchical relations for each module. Every hierarchical relation is expressed, at least in a link in the navigation menu.

 Proximity-based relations can easily be established between any two modules: the author merely has to determine whether the modules are part of the same article, or both report on the same research project, or neither. In the original version of A08, about a third of the references are to other works written by the authors of A05 and A08 or by others in the same group; the rest refer to works by authors outside the project. In the modular version, only 6% of the links express an external relation and 11% a project relation; the rest represent internal relations.

 Range-based relations always occur when a module with a wider range is referred to from a module with a narrower range, notably, when a mesoscopic modules is cited from a microscopic module. In A08, we have provided 39 links to higher-level modules, thus informing the reader of this difference in range. Most of these links appear in the Situation and the constituents of the Methods module.

 Administrative relations are always expressed between each module containing scientific information and the modules providing the meta-information associated to the scientific information in order to allow the reader to navigate between the scientific content of the article and the meta-information. In A08, almost in almost every5.22 content module, three standard administrative relations are made explicit in the navigation menu, in links to the complex module Meta-information, the map of contents and the Abstract. The administrative labels can only be assigned to internal links, which connect modules within the same article.

 Readers of a modular article are forced to take an active part, as they must decide in which sequence they want to consult which modules. This implies that for each link they must decide whether to follow it or not. The reader is less likely to get lost if he has at his disposal a sequential path. Creating these routes is certainly feasible. A default complete sequential route can be generated automatically. However, the author can optimise the route by choosing the most suitable sequence of similar constituent modules of, for example, the Treated results. We have provided the reader with easy-to-use connections to the next module and the previous one by expressing the sequential relations associated with two sequential routes.

 All modules are part of the complete sequential route. Thus, the complete sequential route can guarantee the readers that they haven't missed a single module. However, on this route the readability has been sacrificed for completeness. In the first place, the overlap between different modules (in particular with the module summaries in the complex modules) causes unnecessary repetition for the readers who consult the article via the complete sequential route. In the second place, the bothersome modules like the raw data obstruct the flow of the scientific discourse.  

 The essay-type sequential route is most suitable for consulting the article as a whole. The creation of the essay-type route requires a bit more attention than the creation of the complete sequential route: the resulting route is supposed to provide the sequential reader with a comprehensive account of the research that is represented in the article. In A08, the essay-type route connects all  elementary modules, except the Raw data and the theoretical Treated results. It passes over all complex modules, with the exception of the covering Theoretical methods module A08-m3c and the Quantitative interpretation A08-m5b. These are included in order to facilitate the reader's understanding of the discourse . This is necessary, because both the Theoretical methods and the Quantitative interpretation are very complex modules, which the essay-route visits more than once. If the theoretical methods are used to interpret experimental results, the problem-solving process can be seen as serial: in order to solve the central problem of the article, an experimental sub-problem (of the measurement of, for example, the differential cross sections in chemi-ionisation in collisions between Na and I), and a theoretical problem (of the interpretation of those cross sections in terms of an atom-atom model) are addressed sequentially. This sequence is expressed in the essay-type route, which allows readers to consult the article as a single discourse. In article A08, this route first passes through the Experimental methods and the two constituent modules of the Treated results that deal with the experimental results, before passing through any Theoretical methods modules. The last Theoretical methods module

Both the complete and the essay-type sequential route are part of an article and therefore connect microscopic modules. Therefore, the links are also `internal'. The essay-type route avoids most of the complex modules and therefore the links expressing that route are usually neutral with respect to the hierarchy.

Making explicit the organisational relations gives the reader insight in the organisational coherence of the collection of modules and allows him to navigate efficiently. The organisational types of relations can be identified and expressed in links rather easily. In fact, most of these relations can be made explicit automatically once the modules and the links expressing the relevant scientific content relations are created.  



Next: Representing scientific discourse relations Up: A discussion of the Previous: A discussion of the