||Reflection about the architectural programme starts with the analysis of its writing, its "style" which bears not only the "griffe" of the programmer but as well the structure, methodology, codes of reading, etc. particular to a programming approach. The programme structure corresponds in most cases to the different levels in the text's format and the composition modes of representing data and their relations. The choice made can either facilitate or impede the reading as interpretation of the programme. The programmer’s aim should be to open the text to reading towards a "synthetic schematic" summary, a sort of cognitive threshold which allows the reader to understand both the client's objectives and the designer's intentions enhanced by his experience. Articulating a designer's experience means focusing on his knowhow and memory. The designer's recollected knowledge and heuristic approaches to the solution of a basic design problem - types, his readings and spatial evaluations permanently feed the knowhow. It is important for the architect to have access to past examples, to the collective memory of his workplace, and a repertoire of readings, notes, sketches, influences and citations. It is therfore equally important that a computer environment also have a multimodal "architect's memory" or "project memory" module in which different forms of representation are classified, and made accessible as memory components. It is also necessary to have the possibility to access at any moment in an interactive manner to the recomposition, addition and adaptation of these mnemonic components. The information coming from the programme, classified as descriptive, prescriptive and quantitative types of data, must be able to be interrogated in different modes of representation : text, matrices, nets, diagrams, and so on, so that the pertinent information can be extraded at any given design process stage. Analysis of competition programmes show that often the description of an activity, for example, the Great Stadium competition in Paris, is described by several pages of text, a circulation diagram with arrows and legend, a topological proximity diagram with legend and as table activity - areas . These different representations, which are supposed to be complementary and give the most pertinent view of the client needs, show in fact after analysis, many description problems, incoherance, and which result in a reading difficulty.