hedges

This study aimed to investigate the use of interactional metadiscourse markers in 168 comments made by 28 university students of engineering via an educational forum held as part of a general English course. The students wrote their comments on six topics, with a total of 19,671 words. Their comments during educational discussions were analyzed to determine their use of five metadiscourse categories (hedges, boosters, attitude markers, engagement markers, and self-mentions), making up interactional metadiscourse in Hyland’s (2004) model. Following descriptive analysis of the use of metadiscourse categories, chi-square tests were used to investigate the possible differences in the whole sample as well as gender-based differences. The findings showed that although female EFL learners used more metadiscourse markers than males did, the differences were minor and hence gender did not significantly influence the use of interactional metadiscourse markers. However, while male and female participants used all types of interactional metadiscourse, how they used them varied. They used engagement markers and self-mentions more frequently than boosters, hedges, and attitude markers. Since metadiscourse markers play crucial roles in mediating the relationship between what writers intend to argue and their discourse communities, the results of the present study have obvious importance in increasing students’ awareness of the way they organize their writings.

Hedges, as tools to express tentativeness and doubt, have been studied in plenty of research papers in the Iranian EFL research setting. However, their use in a learner corpus, portraying Iranian learner English, is in need of more research attention. With this end in view, this study aimed at investigating how Iranian EFL learners who have majored in English-related fields in Iran deployed hedges in their academic, expository essays. This study was conducted through running the corpus analysis software MonoConc Pro-Semester version 2.2 on the electronically compiled Iranian Corpus of Learner English, totaling 436,035 words. Automatic and manual analyses suggested that hedges comprised only 7.4% of the total metadiscourse in the Iranian Corpus of Learner English, with 0.68 occurrences per 1,000,000 words. In a comparable native corpus, a sub-corpus of the British Academic Written English, hedges were used with 1.43 occurrences per 1,000,000 words (21% of the total metadiscourse in the corpus). Log-likelihood statistical analysis confirmed statistically significant differences between the two corpora in terms of the use of hedges, with underuse of hedges in the Iranian academic, expository essays relative to the English natives’ essays. Implementations of the results for English academic writing instruction including genre-based, explicit teaching of hedges through data-driven techniques with the aid of tools such as AntConc software and corpora such as the BAWE are considered.