2. RESULTS AND PROPOSALS
Our main results can be summarized as follows:
• Because of the constraints due to the size of the different spaces in which daily life interactions take place, natural knowledge of space is strongly structured into three main representations: microspace (corresponding to the usual prehension relations), mesospace (corresponding to the usual domestic spatial interactions) and macrospace (corresponding to unknown city, maritime or rural spaces...). In consequence, the space representation produced by the usual out-of-school experiences is not naturally homogeneous, and is quite different from elementary geometry.
• In. traditional teaching of geometry, the usual learning of geometric activities requires of pupils many skills from normal microspace relations; these relations are essential to the teaching process.
• In elementary school, when pupils work with geometrical drawings upon paper, then, to save intellectual effort, they naturally use their microspace representation instead of some geometrical knowledge. Consequently, such tasks may hide from teachers some of the obstacles to the learning of basic geometrical concepts; among these are angles, or space as a set of points, or the recognition of certain configurations that arise in geometric theorems.
• Concerning pupils over 12 years old, we have good reasons to expect that one of the main sources of learning difficulties in geometry is the previous treatment of geometrical figures on paper during elementary school. These difficulties come from the way that plane figures are considered in the teaching process, as if they were real objects. When teachers and pupils communicate about drawings on paper, they can show apparent understanding, the former thinking in terms of geometric concepts, the latter in terms of objects.
• We have good reasons to expect that geometric knowledge isn’t spontaneously transferred to solve space problems. This lack of transference generates deficicnces ill the control of spatial abilities, even in low-level professional activities.
• We found how some of the obstacles to learning geometry in the primary school can be overcome by a partial blocking of the influence of visual knowledge of microspace.
Our theoretical analysis makes us pessimistic about the viability of local changes, if these are not linked with essential changes in the space constraints of the pupils' tasks. The need is to prevent the pupils from resorting to some aspects of their personal representation of microspace.
In France, if a teaching process were to take place in the geometry curriculum, which aimed at an improvement of pupils’ spatial knowledge, there is no chance that it would last unless is aims are not explicitly allocated into the curriculum.
If our results were confirmed by other studies, the consequence for the teaching of elementary geometry would be important and ought to be deepened by larger-scale researches.