This article sets out to prove the possibility of a systematic relationship between the syntactic properties, the semantic import, the distribution in discourse and the range of possible interpretations of echo-questions. In particular, it proposes to demonstrate that they are metarepresentations by virtue of their very syntactic structure.