Abstract This article introduces a methodology for the diachronic analysis of large historical corpora, Usage Fluctuation Analysis (UFA). UFA looks at the fluctuation of the usage of a word as observed through collocation. It presupposes neither a commitment to a specific semantic theory, nor that the results will focus solely on semantics. We focus, rather, upon a word’s usage. UFA considers large amounts of evidence about usage, through time, as made available by historical corpora, displaying fluctuation in word usage in the form of a graph. The paper provides guidelines for the interpretation of UFA graphs and provides three short case studies applying the technique to (i) the analysis of the word its and (ii) two words related to social actors, whore and harlot. These case studies relate UFA to prior, labour intensive, corpus and historical analyses. They also highlight the novel observations that the technique affords.

中文翻译：

---

使用波动分析

摘要 本文介绍了一种大型历史语料的历时分析方法，即使用波动分析（UFA）。UFA 着眼于通过搭配观察到的单词用法的波动。它既不假定对特定语义理论的承诺，也不假定结果将只关注语义。相反，我们专注于一个词的用法。UFA 考虑了大量关于使用的证据，随着时间的推移，由历史语料库提供，以图表的形式显示单词使用的波动。该论文提供了解释 UFA 图表的指南，并提供了三个简短的案例研究，该研究将该技术应用于 (i) 单词 its 的分析和 (ii) 与社会演员、妓女和妓女相关的两个单词。这些案例研究将 UFA 与先前的劳动密集型、语料库和历史分析。他们还强调了该技术提供的新颖观察结果。