Physically the result of the project is presented as a book with, attached to the inside of its cover, a diskette with the program, as if that latter ware intended to illustrate some aspects of the former. As far as the contents is concerner, however, the opposite image is more appropriate: that of a program with, attached to it, an accompanying text for the illustration of some aspects of the program.
The most important result of the project is the program Abundantia Verborum (1.0 beta): a ready-to-install Windows program, compiled from some 35000 lines of source code (divided over some 50 unit files), with an online help covering some 200 topics in detail.
By the "beta" in the version number we label the program as ready to be tested thoroughly by the world outside of the local department, rather than having reached the point of being a 100% stable, tested, robust version. At first sight it may seem strange that the presentation of a dissertation coincides with the 'release' of version 1.0 beta rather than version 1.0, but this choice was made in light of the fact that a dissertation by its nature has to be the clearly demarcated work of a single person. For several reasons we found the present to be the very last point in the development cycle for which it is still more or less acceptable that the foregoing has been the plan and work of single person. As the program grows mature, it will inevitable become team work -which mature software is by definition- (a.o. of extensive debugging and documentation), and draw heavily on external tools (e.g. installation software) and external feedback (notably from the beta testing).
Moreover, notwithstanding the beta in the version number the current state of the program is a stable state, as it is the closing point of a phase of internal testing. The early stages of this testing phase had been characterized by a constant need for new development cycles (in terms of the spiral model) to adapt the program to unsatisfied user demands, but that sitation gradually evolved into the current phase of stabilization in which new user questions practically all are covered by the functionality that is already present.
The text of the dissertation serves as introduction to the program. Rather than being a 'document' in its own right, it is merely 'documentation'. Moreover, whereas the dissertation presents a stable point in the evolution of the program, the text of the dissertation is of a far more volatile nature (and is intended as such). In the case of software, constant change would decrease the usability of the program, because one would constantly have to re-learn how to use it (and in the worst case would constantly have version incompatibility problems), so that it is best to only 'release' stable points in the evolution of the program. In the case of the documentation of the program, on the other hand, each new contribution to, or improvement of, the text (new illustrations or example case studies, or additional background information, or corrections of errors, etc.) can only add to the quality of the text. Therefore the text was constructed as a set of HTML-files, intended to be made available on (and easily downloadable from) a web-server. The HTML-files are copied to your system when you install Abundantia Verborum. Updated versions will be available on Locutus, a web-server at the department ( http://locutus.arts.kuleuven.ac.be/ling/). The printer version has been derived from the HTML-version (incidentally, some of the modules of Abundantia Verborum were used for this).
In spite of this volatile nature, however, we wanted the current printed version to contain a coherent story of acceptable size, rather than being a huge collection of heterogeneous background information. Therefore, given the breadth-oriented nature and interdisciplinary aspects of the project, it was necessary to make selections and choose for one main perspective on the subject, out of several possible alternatives, such as:
We judged the third alternative to be the most appropriate for the intended forum for this text, and therefore this has been chosen as main thread. After all, the best way to encourage the use of the program is to write a manual for the software. Actually, we have even taken the line of reasoning one step further and constructed the text in such a way that, for it to be useful as manual, the reader is forced to use the program. We apologize for taking the liberty of doing so. In another respect we did not take the line of reasoning all the way. A very rigid reduction of the text to one perspective would leave many important things unsaid about the design decisions behind the program, and about how the program could be taken one step further. Therefore, we have applied the coherent discourse principle in a flexible way, especially toward the end of the text. Roughly, aspects of the second perspective have found their way to chapter 4. Virtual corpora and corpus linguistics, aspects of the fourth intrude chapter 5. Workshops and corpus linguistics, and, finally, the first perspective pops up in the concluding chapter 6. Conclusions.
This book, the manual to the software, addresses the user of the software. It consists of three major parts. The first part is the purest with respect to our choice for a broad user perspective. In the more advanced topics of parts two and three a strict exclusion of the other perspectives seemed undesirable at some points. Because of the intrinsic interwoven-ness, of the tool with on the one hand, computational aspect of the solutions, and on the other hand, linguistic aspect of the problems, parts two and three contains some more computational and some more theoretical-linguistic passages. We apologize for the points where we were unsuccessful in avoiding abrupt changes of language register.
Apart from these major parts there are appendices covering topics such as a glossary of technical terms, frequently asked questions, exercises for students, example cases studies, etc. Some of these are left out in the printed version, because they currently are hardly more than placeholders for information to come: the indices in particular are the parts of the documentation that are meant to constantly grow and be updated on the basis of new actual use of the program.
As was already mentioned before, there also is an online help to the program. The difference between that document and this text is that the former serves a more technical purpose. Although the online help also contains general background information (so that the program in principle can be used without this manual), it basically is meant for the user who has acquired the basics, but needs a reference guide in which all technical details of Abundantia Verborum version 1.0 beta can be found. The typical use of the online help is either calling context-dependent help from within some dialog box in the program, or looking up some technical term in its index. A related difference with this text is that the online help is more exhaustive than this text, because this text, in its attempt to stay on the main road, fails to see some backyards.