Compression in post-verbal sequences in French

This paper sheds light on the question of whether French resorts to post-focal compression (a reduction of pitch range), and the conditions under which it can do so. We present evidence from a production task in which thirteen native speakers of Standard French read scripted material; i.e. canonical sentences in which we manipulated the type of constituent found in post-verbal position (argument/adjunct), its prosodic length and complexity (short/simple or long/complex) and its informational status (focused/given). The study reveals two main findings: (i) arguments and adjuncts are phrased differently, and (ii) length and information structure exert a significant influence on the realization of “verb + adjunct” sequences. In the sequence “verb + argument”, only a slight compression of the argument is found. We formalize these results by arguing that compression in French is more restricted than in Germanic languages; it can only take place at the level of the prosodic phrase.