A research project is presented in which we aim to design a speech synthesis model based on both the diphone and the allophone concepts, i.e. the data-driven and rule-driven approach for speech synthesis, respectively. At present, diphone concatenation for Dutch leads to more intelligible speech than when a rule-based allophone synthesis is applied, although the latter synthesis has the theoretical advantage of optimal parametric freedom. However, in rule-based speech synthesis the iterative adjustment of the relevant speech parameters appears to be rather laborious. Therefore, an attempt is made to extract information about rules for allophone synthesis from the data-driven diphone synthesis (data-to-rule conversion) by means of a semi-automatic algorithm.