Diachronic changes in lexical density of research article abstracts: A corpus-based study

Lexical density refers to the ratio of content words and measures the information density of academic texts. It has been regarded as an important indicator of writing proficiency and author style in academic writing research. A recent study based on Nature Biology Letters reported an upward trend in lexical density in research articles published in the journal. However, such a finding needs to be validated with a larger dataset. In addition, it remains unknown whether such a trend varies among disciplines. Using a large-scale corpus of research article abstracts from 16 academic disciplines, the present study reveals that the rise in lexical density could be a common trend across all fields. Moreover, the results of cross-disciplinary comparisons show that hard disciplines exhibit higher lexical density than soft disciplines. To our knowledge, this is probably the first study to investigate both the temporal dynamics of and disciplinary variations in lexical density using extensive multidisciplinary text data. Possible reasons for and practical implications of the findings are discussed.