It is argued in this study that research on viewpoint in language use provides better results if the category of voice is recognized as a semi-autonomous construct in relation to viewpoint. Voice is understood here as the “who” in the act of speaking but not identified with the actual speaker. Rather, it is a category of discourse that can be “abstracted”, to some extent, from the broader viewpoint phenomenon. A corpus of journalistic narratives from British quality press and American national press has been analyzed, on the basis of which a theoretical and terminological framework is being proposed. The basic distinction is that between single and multiple voices. As discourse develops, single voices shift from one to another in an abrupt or gradual manner. Multiple voices, in turn, can be “heard” as polyphonic, blended, or fused, depending on the degree of their integration (intertwining), with fused voices resulting in a single collective voice. However, voice is not a fully autonomous category, as it is shown to be closely interlinked with several types of viewpoint. Because this study is based on data from a specific genre of English-language journalism, it is also an invitation to test and expand the framework proposed in analyses of other genres in English and in other languages.

这项研究认为，如果语音的类别被认为是与观点相关的半自治结构，那么对语言使用观点的研究将提供更好的结果。语音在这里被理解为说话者中的“谁”，但并未与实际说话者相提并论。相反，从更广泛的角度来看，这是一种话语类别，在某种程度上可以“抽象化”。分析了来自英国质量新闻界和美国国家新闻界的新闻叙事语料库，并在此基础上提出了理论和术语框架。基本区别是单个声音和多个声音之间的区别。随着话语的发展，单一的声音突然从一种转移到另一种或逐步的方式。反过来，可以将多个声音作为复音，混合或融合“听”到，具体取决于它们的融合程度（交织），融合的声音产生单个的集体声音。但是，语音并不是完全自主的类别，因为它与多种视点紧密关联。因为这项研究是基于特定类型的英语新闻业的数据，所以它也邀请测试和扩展在分析其他类型的英语和其他语言时提出的框架。