1

The problem of homicides during the pre-statistical period in France or in Belgium is addressed by researchers using letters of remission produced through the carrying out of the royal right of pardon. Curves of numbers of remissions have been established from that source and have been considered by some as representatives of actual violence. Indeed, an increase in the number of remissions was interpreted as a sign of a surge in violence. This paper aims at demonstrating that variations of such a curve depend on multiple factors (among which the nearness of the administration and the sovereign) which do not enable to consider remissions for homicides as representative of actual violence. A compared approach between the realms of France and the Low Countries made it possible to highlight connections as well as distinct factors for both spaces.