A recent work by Ramus and colleagues has renewed the interest in rhythm typology connected to the evaluation of speech production data. They proposed that differences in rhythm type could be accounted for by a segmental set of variables derived from the acoustic duration of consonants and vowels. However, rhythm typology can be more interestingly characterized and understood by modeling speech rhythm production. The purpose of our study is then threefold: (a) showing why a deeper understanding of languages¡¯ rhythmic types can only be achieved by modeling their underlying rhythmic systems; (b) presenting the results of the implementation of a coupled-oscillator rhythmic system which simulates language-specific continuous patterns of syllable-sized durations; (c) suggesting that only a rhythmic system integrated to a gestural framework can account for the complexity of rhythm phenomena.