Investigating legal language peculiarities across different types of Italian legal texts: an NLP-based approach

In this paper, the author carried out the linguistic profiling of a corpus of different types of Italian legal texts exemplifying different sub-varieties of Italian legal language by relying on a wide range of different linguistic features (lexical, morpho-syntactic and syntactic) automatically extracted from the output of a multi-level automatic linguistic analysis of texts. The devised comparative approach allowed investigating the linguistic variation i) between the considered corpus of legal texts and a corpus of newspaper articles representative of Italian ordinary language and ii) among the considered types of legal texts (legislative acts, administrative acts, the Italian Constitution and legal cases). Achieved results can provide the starting point to identify areas of lexical, morpho-syntactic and/or syntactic complexity within a legal text in order to assess its readability as well to perform a number of different computational forensic linguistics tasks.