Abstract Informed by image repair theory, this study examines grammatical resources used to represent agents of retraction-engendering acts in retraction notices (RNs). A corpus of 250 RNs from two broad disciplinary groupings and authored by different stakeholders was analyzed to determine if agents of retraction-engendering acts were identified and how linguistically visible they were made. It was found that agents of culpable acts were identified in only 44.40% of the RNs and that agent-obscuring grammatical resources were deployed about 3.35 times more frequently than agent-identifying ones were. Furthermore, the hard-discipline RNs authored by journal authorities identified agents of culpable acts significantly less frequently and less explicitly than both the hard-discipline RNs from authors of retracted publications and the soft-discipline RNs written by journal authorities did. These results suggest that choices of grammatical resources in RNs are influenced by a complex web of factors, including different retraction stakeholders’ varied communicative purposes, their image repair efforts, their relation to the reprehensible acts, and legal considerations. These findings warrant further attention to language use in RNs as a high-stakes genre.

中文翻译：

---

代理和责任：撤回通知中的过错行为的语言分析

摘要 在图像修复理论的指导下，本研究检查了用于表示撤回通知 (RN) 中引起撤回行为的代理的语法资源。分析了来自两个广泛学科组并由不同利益相关者编写的 250 名 RN 的语料库，以确定是否识别了导致撤回行为的代理人，以及他们在语言上的可见性。结果发现，只有 44.40% 的 RN 发现了有罪行为的代理，并且代理模糊语法资源的部署频率是代理识别语法资源的 3.35 倍。此外，与来自被撤稿出版物作者的严格 RN 和由期刊当局撰写的软纪律 RN 相比，由期刊当局撰写的严格纪律 RN 识别有罪行为的代理人的频率和明确程度要低得多。这些结果表明，RN 中语法资源的选择受到复杂因素网络的影响，包括不同收回利益相关者的不同交流目的、他们的形象修复工作、他们与应受谴责的行为的关系以及法律考虑。这些发现值得进一步关注 RN 中的语言使用作为一种高风险类型。包括不同收回利益相关者的不同交流目的、他们的形象修复工作、他们与应受谴责的行为的关系以及法律考虑。这些发现值得进一步关注 RN 中的语言使用作为一种高风险类型。包括不同收回利益相关者的不同交流目的、他们的形象修复工作、他们与应受谴责的行为的关系以及法律考虑。这些发现值得进一步关注 RN 中的语言使用作为一种高风险类型。