Ten years of cognitive dimensions in visual languages and computing: Guest Editor's introduction to special issue

This paper marks the tenth anniversary of the 1996 publication, in the Journal of Visual Languages and Computing, of Thomas Green and Marian Petre’s classic paper Usability analysis of visual programming environments: a ‘cognitive dimensions’ framework [1]. In the ten years since then, many visual language researchers have used this paper as their primary reference with regard to usability questions, and it has spawned a large secondary literature, with over 50 publications devoted to the further development of the cognitive dimensions (CDs) framework. The papers in this special issue represent the current state of the art in the field. At the time of writing, the Green and Petre paper is the most-cited original research publication that has appeared in this journal (another highly cited paper is the taxonomic review in [2]). This issue builds on the subsequent interest in usability of visual languages, with a particular emphasis on the ways in which the framework is becoming relevant in professional design, in characterisation of language usability, and in extension of visual languages approaches to other recent technology developments. Green initially proposed the cognitive dimensions approach in the late 1980s [3,4]. His motivation was better to incorporate research findings from applied psychology into research and practice in technology design, an issue that remains a critical concern in the field of human-computer interaction today. Petre’s own investigations into skilled professional practice in contemporary engineering work brought particular emphasis on the application of the approach to understanding visual design notations. The publication in this journal of the resulting collaboration coincided with increasing concern for improving the usability of visual languages, as evidenced by Green’s invitation (with Ben Shneiderman) to give opening and closing keynote addresses at the IEEE Symposium on Visual Languages in Boulder, Colorado in 1996. Before that time, research into novel visual languages had already been motivated by the desire to create languages that were either accessible to a wider range of users, more effective for expert users, or offered both types of benefit. Early research in the field was able to proceed on the basis of intuitive preferences for particular representation styles, but increasing diversity of visual language designs during the 1980s and 90s resulted in the need for some means to compare the relative benefits to users of alternative languages. One option was to use experimental methods from applied psychology, as Green himself had