Certain confusion may be observed in the field of biomimetic architecture, as it emerges at the crossroad of two disciplinary domains: architectural design and biological sciences. If biomimetics is defined as a science, once it is applied to architecture, biomimetic architecture should logically be defined as a science too. This assertion collides with the nature of architectural design, which may rather be defined as a technology: its aim is to transform the world, not to explain it. On the one hand, there is no obvious relationship between architecture and life sciences. On the other hand, the biomimetic approach tends to redefine the concept of science itself by seeking to avoid the excesses of scientism. Moreover, existing applications of biomimetic design show that it is difficult to observe a genuine biomimetic architecture; most cases are closer to engineering component or urban planning and sometimes they involve little or no life sciences. The aim of this paper is to describe this conceptual confusion through two movements called “forcings”, occurring between design and science. These forcings are conceptualised as shifts between constructed scientific objects and given empirical objects. Models, used in biology as in architecture, allow these shifts by virtue of their double function. They are both tools for knowledge and for design, thus they may be conceptually forced into what they are not supposed to be, particularly in the field of biomimetic architecture where design process and scientific knowledge are said to meet.