The main purpose of the academic writing is to inform other researchers about writers’ findings in certain research. In this case, writer will propose claims. For non-native English speaker like Indonesian, this is the tough work to do. L2 learners find difficulty to write for academic purposes or make claims. One of the strategies that L2 learners do is by using hedging devices. Hedges are used to present findings cautiously with leaving room for readers to have their own interpretation. This argument is also supported by Ken Hyland (1996) stated that academic writing is full of hedges. This study aims to find the hedges in academic writing used by Indonesian researchers or writers. According to Levinson (1987) with his theory of FTA (Face Treathening Act), those words mostly function as a tool for speakers or writers to make them comfortable and save negative face. It means that the writers should choose the correct words to achieve the communicative goal. The data is taken from 10 dissertations written in English. The method used is decriptive-qualitative analysis. The study focuses on 2 kinds of hedging strategies proposed by Hyland (1996). They are writer-oriented hedges and reader-oriented hedges. The first strategy consists of (1) passive voice, (2) dummy subjects, and (3) abstract rhetors. The latter consists of (1) personal attribution and (2) conditionals. The results reveal that writer-oriented hedges are the most frequent hedging device utilized by Indonesian researchers, such as: passive construction and dummy subjects. The conclusion of this study is that the use of passive constructions and modality (can, may, might, should) are highly desireable by Indonesian researchers. It means that Indonesians like to let the data talk by themselves in order to avoid a potential conflict and hence to maintain the harmony between writers and readers.