Abstract:

With an impressive development of science and technology over past decades, more scientific papers are published now before ever before. At the meanwhile, owing to the expansion of scientific research field, it is increasingly difficult to get research papers published in high impact journals, which is very important to scientists since the quantity and quality of publications have a considerable effect on their reputation and promotion. Consequently, in order to get papers published, scientists had to change their scientific writing style, which led to the evolution of scientific writing style over time. In this paper, we try to find the potential patterns under the evolution of scientific writing style between 1994 and 2012 from the emotional vocabulary perspective. The dataset was selected from China Journal Fulltext Database and Association for Computing Machinery(ACM) Fulltext Database, which contains all 215,515 scientific papers from “computer science” in Chinese, and 161,428 from “computer science” in English. This experiment was conducted mainly from the following several aspects. First of all, we selected 75 emotional vocabularies (25 positive words,25 negative words and 25 neutral words included) by three times group brain storm. Second, we depict the yearly absolute and relative frequency patterns of positive word, negative word, neutral word and 100 random words respectively between 19942012. Finally, we compared the absolute and relative frequency patterns of emotional vocabularies in different language and different level of journals. By the comparative analysis with the absolute and relative frequency patterns of emotional vocabularies, we find that the emotional vocabulary is both increasingly used in scientific writing in different languages and all levels of journals, especially positive words, which probably related to the emergence of positive scientific findings and also reflects scientist are more likely to use an optimistic writing style.

Key words: Writing style, Emotional vocabulary, Data driven, Retrospective analysis, Scientific papers, Evolution