The ambiguity apparent in the definition of task can also be detected when considering the way tasks appear in practice in the classroom: in the form of different task types. Based on a review of relevant literature, it may be concluded that there seems to be no consensus concerning the exact number and the precise nature of existing task types. While traditional task type labels originating from the classroom, such as information gap, jigsaw, opinion exchange, decision making, or problem solving, also frequently occur in articles, the reader may easily get confused by phrases such as one-way and two-way tasks (Long, 1985), convergent and divergent tasks (Duff, 1986), or static, dynamic, and abstract tasks (Brown, Anderson, Shilcock & Yule, 1984). These latter task classifications are all based on certain distinctive task features the authors judge to be of crucial importance with regard to the pedagogical usefulness or the difficulty of tasks, which are as follows: the direction of information flow between those interacting, the number of solutions that can be arrived at, and the information type the task contains, respectively. Here the terminological diversity results from the fact that different researchers emphasise different aspects of tasks. The issue is complicated further by the varying levels of analysis employed and the different levels of specificity involved: the traditional task types, jigsaw and information gap, can also be classified as convergent tasks; the Spot the difference (Plough & Gass, 1993), Draw the picture (Gass & Varonis, 1985) and Assemble the scene (Pica, Young & Doughty, 1987) tasks are all information gap tasks but have different content.