Writing with suspense: A corpus-based news lead analysis

News lead is regarded as one, if not the most, important part of news articles since it bridges the gap between headlines and the content. It is the first paragraph, or the first string of sentences, which comes to readers’ view when they start reading. Therefore, attractive leads can ‘hook’ readers and keep them curious as the news story unfolds. Journalists and editors, in view of the significance of a ‘good lead’, are trained to write leads as interesting as possible. The drilling is deemed essential to many journalism students that celebrated textbooks like Writing and reporting for the media (Bender, Davenport, Drager & Fedler, 2016) specifically teach lead-writing skills in extensive chapters. However, many coaching guides only introduce various types of leads but fail to systematically present the syntactic features and semantic categories which are marked in different news types. While traditional media practitioners prefer ‘summary leads’, alternative ones such as ‘buried leads’ and ‘quotation leads’ are popular in feature stories. Journalism students, like my old self, often have difficulties in deciding which style suits their stories most, or what semantic elements are required in a certain lead type. To facilitate undergraduate news writing courses, my project aims to fill in the said gap to provide ESP resources for journalism students to enhance news lead writing. By drawing on leads from authentic articles of wire news, hard news and feature news respectively, a ‘news lead’ corpus is built and automatically parsed to study the syntactic structures favoured in each sub-corpus of news type. The corpus is then annotated with the UCREL Semantic Analysis System (USAS) tagger to summarise the semantic categories preferred by each news type. The two sets of findings, driven from mass empirical data, can complement the existing guidelines to help budding journalists write an appropriate lead.