Sentences with a local syntactic ambiguity were created (e.g. '' [le petit chien mort]...'' - '' [the little dead dog]...'' - for the adjective target versus '' [le petit chien] [mord...] '' - '' [the little dog] [bites...] - for the verb target ; brackets indicate phonological phrase boundaries). There were recorded by an expert speaker either with an informative prosody (very salient acoustic variations) or a more neutral one (cues largely minimized). In both a completion task and a word detection task, they found that French adults distinguished between two sentence beginnings that differed only syntactically and prosodically. Before they had access to the disambiguating information, adults gave more adjective responses to adjective sentences than to verb sentences, and vice-versa for verb responses. In addition, ambiguity resolution was better when prosody was more salient.
In order to test whether naïve speakers spontaneously produce these prosodic cues, and if they guide the resolution of temporary syntactic ambiguities, sentences similar to Millotte et al.'s stimuli were read by six speakers (intermixed with control sentences without any ambiguity). All speakers were unaware of the aim of the experiment and they did not realize that half of the sentences were ambiguous. These sentences were cut just after the ambiguous word and presented to French adults in a completion task, in which participants listened to the beginnings of ambiguous sentences and completed them in writing. Results were similar to those obtained by Millotte et al. in the informative prosody condition: participants more often interpreted the ambiguous word as an adjective when they heard adjective sentences than when they heard verb sentences (and vice-versa for verb interpretation). This result indicates that naïve and untrained speakers spontaneously produce phonological phrase boundary cues that are salient enough to constrain listeners' syntactic analysis of spoken sentences.