We propose to represent the syntax and the semantics of natural languages with labelled proof nets inside the implicative fragment of intuitionistic linear logic. Resource-sensitivity of linear logic is used to express all kind of dependencies between the syntagms of a sentence under the form of a proof net. Phonetic and semantic labelling of the proof net from its inputs to the unique output are used to produce the well-formed phonetic form and the semantic representation of the sentence from entries of a lexicon.