In contrast to past consensus, many authors now feel that the passive voice compromises the quality of scientific writing. However, studies involving scientific articles are rare. Using a corpus of 60 scientific research articles from six journals, this study examined the proportion of passives used, and the contexts and forms in which they occurred. The results revealed that about 30% of all clauses were passive clauses. The canonical form was most pervasive, followed by the bare passive; together, they constituted more than a quarter of all clauses analyzed. Passives were typically used in main clauses, followed by relative and adverbial clauses. Roughly 29% of all passives were located in the methodology section. Based on the results, the proportion of passives in scientific writing may stabilize at about 30%. It is unlikely to dramatically drop any further since the trend suggests that passives are still widely used in the methodology section.
KEYWORDS: Scholarly publishing, science writing 1 Context and objectivesA cursory glance at the numerous writing guides available, both on the Internet and in printed form, reveals a clear shift in consensus regarding the use of the passive voice in scientific writing. While the prevalent use of the passive voice was characteristic of scientific discourse in the major part of the 20th century [1,2], the situation today is rather different. Many writing guides favor the use of the active voice for reasons of clarity and conciseness [3]. In their writing guide for scientific English, Day and Sakaduski [4] state this preference even more directly -"the passive voice should be avoided".Yet, as Ding [1] has shown in his work involving corpora from the 18th century to the end of the 19th century, early scientific discourse actually favored the active voice. The move towards the passive voice in the 20th century arose as a result of the increasing demands for scientific discourse to be objective, in the sense that the writing should ideally represent the world "in terms of objects, things, and materials" [1] rather than humans. As a consequence, scientific writing took on a decidedly object-or thing-oriented character [5,6]. With the preference for the active voice in present-day scientific writing, we appear to have come full circle.Indications from an informal survey among postgraduates at a Singaporean university, however, suggest that not many may be aware of this current preference. In fact, 90 of JCOM 13(01)(2014)A03