摘要：This paper describes the process of annotating pronominal anaphor in a corpus of Basque which consists of 54.000 words. Our aim is to use this annotation as a basis for later computational processing. The linguistic study carried out and the criteria defined for the tagging process are also presented in the paper.
