The Mandarin, or standard Chinese has become the center of all attention in automatic language processing. From a scientific point of view, the quantity of data produced in this language, especially via the new means of communication, makes it a favored playground for automatic language processing (LAL) and in particular automatic learning. In this article, we review the main tasks of the TAL and point out the difficulties specific to the automatic processing of mandarin. The latter are intimately linked to the linguistic peculiarities of this language. We hope to encourage dialogue between linguist communities and Mandarin NLP specialists.