Abstract This paper reports on a corpus-based study of linguistic expressions of surprise (i.e., a type of attitude markers functioning as metadiscourse) in 160 applied linguistics research articles that were published in two periods of time separated by 30 years. Unlike previous research on metadiscourse, this study took a frame semantics perspective on surprise as a knowledge emotion and adopted a fine-grained, frame-based analytical framework to examine diachronic trends in the use of surprise markers and their co-occurrence with other types of metadiscourse (i.e., boosters, hedges, and self-mentions). Binary logistic regressions revealed that compared with research articles published earlier, those published more recently were 2.16 times more likely to express surprises triggered by prior knowledge, 2.37 times more likely to express surprises without providing an explanation, and 2.28 times more likely to indicate surprises without resolving them. These results can be explained in terms of the heuristic nature of surprise and the escalating pressure on academics to promote their research strategically.

中文翻译：

---

应用语言学研究文章中的惊喜标记：历时视角

摘要 本文报道了在相隔 30 年的两个时间段内发表的 160 篇应用语言学研究文章中基于语料库的惊奇语言表达研究（即作为元话语的一种态度标记）。与以往对元话语的研究不同，本研究将惊喜作为一种知识情感的框架语义视角，并采用细粒度的、基于框架的分析框架来检验惊喜标记使用的历时趋势及其与其他类型的共现。元话语（即助推器、对冲和自我提及）。二元逻辑回归显示，与较早发表的研究文章相比，最近发表的研究文章表达由先验知识引发的惊喜的可能性高 2.16 倍，2。在不提供解释的情况下表达惊讶的可能性高 37 倍，而在不解决问题的情况下表示惊讶的可能性高 2.28 倍。这些结果可以从惊喜的启发性本质以及学术界战略性地促进他们研究的压力不断增加来解释。