|Table of Contents|
|Table of Contents|
The modular model is meant for the creation and the evaluation of new modular scientific articles, not for the recreation of published articles. In this work, however, the modular model is used to create modular versions of (or in other words to `modularise') published articles, because we do not aim to write a particular modular article, but to develop and evaluate of a modular model, and modularising linear articles allows for the comparison of a linear version with a modular version of the same article.
We chose to modularise published articles in order to check if our modular model is compatible with scientific practice. The original versions of the published articles can be used as a bench-mark, because their publication in refereed journals supplied them with a mark of approval.4.1 The established format of present-day scientific articles is, as we mentioned in section 2.1.3, the product of a long evolution, in which it has been adapted to the needs of the scientific community. We have taken that established format as a starting point in the development of the modular model.4.2
We have at our disposal a corpus of publications on a coherent and complete research project in the field of experimental molecular dynamics. That research has been carried out by prof.dr. J. Los and co-workers at the FOM-Institute for Atomic and Molecular Physics in Amsterdam (AMOLF) between 1968 and 1986. The corpus is described in section 5.1.1 and the bibliography of the publications it comprises is given in appendix B.
The choice of a coherent corpus on a single research project allows for the analysis of not only individual articles, but also of the connections between the different articles. In a coherent corpus, the network of information represented in one article can be explored, as well as its embedding within the larger network of all available information. The restriction to a single research project, conducted at a particular laboratory and during a particular period of time, may limit the validity of our findings. We assume, however, that the research reported in our corpus is typical of its domain.
This particular research project has been selected because of its internationally recognised quality and its variety: experimental, theoretical and numerical aspects of the subject are studied. It has led to publications, which are widely cited, in different international journals, conference proceedings and theses. The publications conform to the standard practice of scientific publication, following the style favoured in the established journals. The fact that the senior author and project leader has been willing and able to assist us with our analysis also played a role in the decision to analyse a corpus on this particular research project.
We have chosen a corpus on the subject of experimental molecular dynamics for the following reasons. Firstly, experimental molecular dynamics, which forms a bridge between physics and chemistry, is a rather prototypical domain in science. It is an established field of `normal', problem-solving science (see section 2.1.2). Secondly, publications in this domain are relatively accessible; no specialised mathematical skills are required for a basic understanding of the subject. Thirdly, whereas on the one hand articles on experimental molecular dynamics are simple enough to allow for a comprehensive analysis, on the other hand experimental, theoretical and numerical aspects are intertwined in the problem-solution pattern, so that the structure of the research in the domain, and thereby of the articles about it, is complex enough for the analysis to be non-trivial.
So, a corpus has been chosen that is coherent, accessible and typical of publications in scientific practice.