Hedging in hard and soft disciplines: A corpus-based analysis of research articles

The article is focused on a quantitative analysis of the use of hedges in a 1.6-million-word corpus of academic articles in four hard sciences (chemistry, physics, mathematics, engineering) and four soft sciences (linguistics, history, management, political science). It is hypothesised that there are significant differences in the use of hedges between the two groups of disciplines. The analysis is based on an examination of 81 language units which previous studies have found to be frequent in academic discourse. These language units were divided into six groups based on their morphosyntactic features: modal verbs, lexical verbs, noun phrases, prepositional phrases, adjectives and adverbs. Statistical modelling of the data confirmed the hypothesis, revealing significant differences in the use of all the considered types of hedges in hard and soft science articles. It was also shown that the types of hedges such as adverbs, lexical verbs and noun phrases contributed the most to the interdisciplinary differences. All these hedges are more frequent in soft sciences. The application of the hierarchical clustering algorithm revealed two stable clusters that correspond almost entirely to the two categories of disciplines under study (with the exception of mathematics and political science), thus demonstrating that hedges can be used as a means of differentiation between the two groups of sciences.