We often tackle the territorial problem only with a regressive analysis. However this reasoning tends to enhance the status of settlements still seen nowadays as villages or towns. On the contrary, we think that protohistoric agglomerations were much less significant if they were deserted during gallo-roman period. That's why our reasoning is influenced by our visible cultural heritage. Yet a spatial approach about the dynamic of settlement systems during the Iron Age underlined a non-linear development with relative "recession" and "expansion" periods for each agglomeration. Starting with the fifth century BC, these compete in establishing their (extractive) settlement networks, and in developing their land use. We can observe in a relatively detailed way the intensification of this competition in the form of symbolic expressions of the networks' growth through the visibility control, the increasingly massive fortifications and the increasingly high towers. Extern stimuli, as the Mediterranean exchanges or the roman colonisation for example, seem to accentuate this competition and contribute to develop some agglomerations or to pressurize others. We propose an analysis model which cross both reasonings, regressive and spatial that would give a better understanding of the evolution of territorial patterns during Iron Age and what it brought in our territorial heritage.