We make study models constantly, from the beginning to the end of the project. The scale gradually becomes larger, but whether we are going forward or taking a few steps back, it’s something that we repeat again and again as we create the real thing. We build a large number of models at the beginning of the process. But once the basic design phase is over and we move on to the implementation process, we make one good model at about 1:50 or 1:20, and instead of setting it aside for another, as we had done up to that moment, we use it until the end, modifying it as necessary. At the same time, we use the small models from a slightly earlier stage to check the overall design, while building any number of partial models on a larger scale to pull together our thoughts about details. The difficult thing about study models is how to use them to think about reality with the right level of abstraction. Or, to put it the opposite way, when you think about architecture, when you build it, at some point there has to be a process of abstraction. I think making models is useful for that. When the materials are very important, or when it is a special kind of structure, then something similar has to happen in the model...[+]