Writing for computer science.

From MaRDI portal





[For a review of the first edition (1997) see Zbl 0882.68151.] When the editors of Zbl entrusted me with the review of Zobel's book, my first reaction was: why should anyone write about how to write about research results in computer science? In which way could that be different from good writing about good scientific work in any of the sciences? I therefore decided to put the book aside for a while, and only later formulate a brief neutral or ironic comment. However, after now having read most of the book and skimmed the rest, I decided to recommend Zobel's work to my advanced students and Ph.D. candidates. No student will become a researcher, or a technical writer, just by reading a book about how to conduct research, or how to write a paper. But those who are motivated, or even driven, to do science, and have done their first steps into the exciting world of conjectures, hypotheses, experiments, implementations, proofs, and refutations, or defeats may take advantage of some of the advice contained in the chapters of this volume. Zobel's book is charming because his way of writing conveys a great degree of dedication to science. There is a large number of ``How to books in the area of degrees, seminar papers, or theses work. Some are fun to read, some are outrageously dull, and all of them tend to promise more than they could possibly satisfy. Zobel manages to collect a great deal of immediately useful practical hints, and to discuss a number of processes that are of relevance to research projects, but are hard to deal with by trying to give advice. Zobel's general attitude towards science is that you must take your time, always be skeptical and open, grasp each opportunity to communicate, and be precise and consistent. Only the latter is the area where general advice may be given in the form of rules of experience. And even here, Zobel warns his readers: ``By all means break a rule, but have a good reason for doing so. (p. 5) Five of the fourteen chapters are on general aspects of good writing style (``Good style, ``Style specifics, ``Punctuation, ``Editing, ``Writing up). They contain much of what you should know but often neglect with bad excuse. Nothing of that is specific to computer science. Three chapters are on various components and elements that would appear in many computer science papers or other publications: mathematics, graphs, figures, tables, and algorithms. There are, of course, advanced books on each of these topics, in particular on the visuals (Zobel refers to Tufte's great work here). Even the presentation of algorithms has been dicussed a bit, but this topic appears as the most specific issue when it comes to writing for computer science. Four more chapters treat matters of organizing one's research, including issues of an ethically responsible author (``Doing research, ``Experimentation, ``Refereeing as a surprise, ``Ethics). The last chapter briefly introduces questions of presenting research results in a talk. It could have been omitted. But if we take the Ph.D. candidate as the main reader, it is a good idea to give him or her some final advice on how to prepare for the thesis presentation. An extra goody are the twenty exercises. They constitute a good start into raising many of the issues treated in the book. Many chapters give check-lists or lists of good practice. All of them contain positive and negative examples of style, often in form of pairs that clearly show the improvement. Zobel's book never makes readers believe that they could become good writers by mechanically following general rules. His own way of writing demonstrates that there are helpful rules no writer should disobey but that -- with good reason -- could always be broken. Good writing, even in computing science, remains a process of experience and dialectics. Zobel's book proves the point.











This page was built for publication: Writing for computer science.

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4818645)