Is the high plateau in a ‘hat pattern’ a phonetic artefact, or does it reflect a phonological feature? Can it contrast with a low plateau, i.e. a ‘dip pattern’? The presented perception experiment supports the phonological point of view, since it shows that the dip/hat contrast can disambiguate German oder-constructions, which are interpretable as ‘alternative’ or ‘yes/no-questions’. This specific function may be derived from a more general substance–function relation: While a hat pattern has a ‘bracketing function’, a dip signals detachment.