本文是對職業譯員英中無稿同步口譯停頓頻次特徵的個案研究。研究者選取VOA（美國之音）職業譯員對美國總統川普（Donald Trump）英文就職演講的現場英中直播同步口譯為個案語料，建立了小型雙語雙模態平行語料庫，並以0.250秒為停頓時長闕值對於演講原文和同步口譯譯文中出現的所有停頓都進行了標注。使用SPSS社會學統計軟體將原文與譯文中停頓頻次加以比較並進行卡方檢驗，結果顯示，譯文中的停頓頻次要超過原文，p<0.05，差異具有顯著性。針對量化分析結果，研究者進一步對原文和譯文中的停頓在文本模態和音訊模態下分別進行深度質性語料分析，發現了五種導致英譯中同步口譯譯文停頓相比原文增加的五種操作模式：（1）緊密語言結構的鬆散化；（2）長訊息結構的切分；（3）譯文的分批次產出；（4）與副語言特徵的共現；（5）交際效果的強化。本研究結果有助於深化瞭解同步口譯認知過程，對於同步口譯教學和實踐也有指導意義。

This paper is a case study of the pause frequency characteristics of professional interpreters’ English-Chinese (EC) simultaneous interpreting without text. The VOA live EC simultaneous interpreting of Donald Trump’s presidential inaugural speech is selected as the case, based on which a small-size bilingual parallel corpus with double modalities is built. A comparison of pause frequency and its chi-squared test is made through SPSS between original speech and the EC simultaneous interpretation, whose result shows that the pause frequency in simultaneous interpretation is bigger than that in original speech. The above quantitative analysis is followed by an in-depth qualitative one under both audio and text modalities concerning those pauses in the corpus, the result of which finds five operation modes as follows causing pause frequency increase in Chinese interpretation: 1) loosening of compact structures; 2) segmentation of long information structures; 3) production of interpretation in batches; 4) co-occurrence with paralinguistic features; 5) strengthening of communicative effects. Findings of this research may provide some insights for teaching and practice of simultaneous interpreting while adding to the knowledge of its cognitive process.