RSS feed

Coping Strategies for Fast Delivery in Simultaneous Interpretation

Changshuan Li, Beijing Foreign Studies University


Fast speech is the arch enemy of simultaneous interpreters. Prior preparation may address deficiencies in knowledge and terminology, and to some extent, alleviate the pressure of speed of delivery. But if the speed is beyond a certain limit, no interpreter can transmit the message in full, even if he/she is an expert in the subject. Fast deliveries place interpreters and listeners, particularly non-native listeners, at a disadvantage. It is widely recognised that a rate between 100 and 120 words per minute (wpm) is optimal for English speeches, although the figure may differ for different speech types. This translates into an optimal speed of 150-180 syllables per minute for Chinese speeches. To cope with speeches faster than optimal speeds, this paper proposes four strategies: the speaker is requested to slow down; the interpreter speeds up; summarisation; termination of service.


Simultaneous interpreting, delivery speed, coping strategies, summarisation, termination of service.

1. Speed—an Insurmountable Barrier

Fast speech is the arch enemy of simultaneous interpreters. Prior preparation may address deficiencies in knowledge and terminology, and to some extent, alleviate the pressure of fast delivery. But if the speed is beyond a certain limit, no interpreter can transmit the message in full, even if the interpreter is an expert in the subject. This is even more so when the structures of the source and target languages differ substantially. The reason is simple. All interpreters have limited mental capacity. In simultaneous interpretation (SI), interpreters have to allocate attention among several tasks: listening and analysis, production, short-term memory and coordination (Gile 1995: 161). When a speaker’s delivery is rapid, listening and analysis alone will consume almost all the interpreter’s energy. Little energy will be left for production, especially when production involves complicated language restructuring.

The human brain is like a washing machine. The drum must never be overloaded with laundry, or there will be no room for spinning, and cleaning will not be thorough. Likewise, fast speeches overload the brain with too much information within a specific time span, leaving no room for proper processing of information to produce a coherent translation.

2. Problems of Fast Speech

Delivering fast speeches in an international conference may lead to several problems: First, mistranslation and loss of information. There is a maximum output that an interpreter can produce within a given time interval; the greater the input, the greater the chance of error and omission. Secondly, fast delivery makes comprehension difficult even when the audience is listening to a native language. Thirdly, English is a non-native language to many, if not most, international conference participants. Participants lose information when either the speaker or the interpreter speaks too fast.

On speed of delivery vs. comprehension, Xie Likui of Hubei Radio Station in China argues that there is a limit to speed in language production and language reception. When a speech is delivered too fast, ‘the ears’ cannot catch up with ‘the mouth,’ and the listener has difficulty understanding the speaker. According to Xie, the speed of news broadcasting has become faster. Take News Digest (Xinwen he baozhi zhaiyao), a China Central Television (or CCTV, China’s national TV network) programme as an example. In the 1960’s, the broadcaster’s speed of delivery was 185 characters per minute (cpm) (each character is one syllable). The speed was increased to 200-220 cpm in the 1980’s, 240-260 cpm in the 1990s, 250-270 cpm in recent years, and in extreme cases, over 300 cpm today. Speech spoken with speed taxes the listener and hampers understanding. In a letter to CCTV, a listener complained: “News Digest broadcasters read the news incredibly fast. They speak like rolling thunder. I can hardly breathe when I listen” (Xie 2002). When speeches are spoken so fast to the extent of causing breathing difficulties, reception and comprehension will suffer.

3. What is the optimal speech rate for English?

To ensure the proper functioning of the interpreter’s brain, the speaker must speak at an appropriate speed. Studies show that speech rate has a direct correlation with interpretation quality. Accuracy is reduced as the speech gets faster. Interpreters generally believe that, to ensure the quality of interpretation, a rate between 100 and 120 (English) words per minute (wpm) is optimal for speeches that are not read from a written text, although the figure may differ for different types of speech. Lederer suggested that for recited texts which are devoid of hesitation and redundancy that characterise official speeches, the maximum rate should be 100 wpm (Gerver 1969; Seleskovitch 1978; Lederer 1981; as cited in Chang 2005: 12).

The results of various studies concur with the general beliefs of simultaneous interpreters. The following is a comment taken from the message board of the International Association of Conference Interpreters (AIIC):

We all know that a speaker speaking at about 100 to 120 words per minute is perfectly acceptable. However, there are exceptions to this: dense originals without much redundancy. Such speeches may seem excessively fast even if presented at 120 words per minute. Normally, however, any original exceeding 140 words per minute is fast. I have clocked speakers at 180 words per minute (Communicate 1999).

For purposes of comparison, the speed of delivery of Voice of America’s (VOA) Special English programme is 90 wpm, or two-thirds of normal broadcasting speed (140 wpm) (VOA 2005). In China, the broadcasters of China Radio International English programmes speak at a rate of 120 wpm, which is still fairly fast to Chinese ears.

4. What is the optimal speech rate for Chinese?

As in any other language, a speech in Chinese must be delivered more slowly than reading news if interpretation is to be effective.

According to Professor Zhou Tongchun, a phonetics expert at Beijing Normal University, if the language used is everyday language, the human ear has a capacity of receiving seven to eight characters per second, or 240 to 250 cpm. Beyond this speed, the listener has difficulty in comprehension and perception. Xie Likui noted that (Chinese) listeners have a recognition rate of 240 to 250 cpm, and as high as 280 cpm, if broadcasting is in Putonghua (Mandarin); hence, based on the principle of serving the majority, a broadcasting speed of between 250 and 260 cpm is more appropriate (Xie 2002).

Is this speed appropriate for simultaneous interpretation? The author did a word count on the number of words in the English version of The Government’s Work Report (Zhengfu gongzuo baogao) that Premier Wen Jiabao delivered at the 2009 National People’s Congress. The ratio of the number of Chinese characters (syllables) to English words is approximately 1.3:1, meaning that every 1.3 Chinese character is translated into one English word (Note: based on the author’s experience, with succinct writing, depending on the nature of the document, this ratio may be as high as 1.7:1).1 Using the broadcaster’s speed of 260 cpm as a benchmark, if the English text was read aloud simultaneously, the reading speed (i.e. “shadowing,” without any processing) will have to be 200 wpm to keep pace. This is a speed that greatly exceeds the speed of English broadcasters (the reading speed of Standard English rendered by BBC and VOA broadcasters is about 140 wpm). Even if the interpreter manages to process and deliver the information, few among the audience will understand. Fortunately, the annual Government’s Work Report is always read by the Prime Minister of the country (not a news reader), who is usually a person speaking at a measured pace. Feng Jingbao, an experienced interpreter, said: “If a speaker speaks at the speed of the CCTV news reader, no one is capable of interpreting” (Feng 2002). Therefore, SI from Chinese to English is impossible for speeches delivered at the speed of normal news broadcasting. Based on the author’s experience, when a Chinese speaker reads from a prepared text at a fast speed, the interpreter will not be able to catch up with the speaker even if he/she were reading from a prepared English translation, unless the translation is exceptionally succinct.

The author has performed no empirical study to find out the optimal speed for Chinese delivery. But a simple calculation could provide some indication. As mentioned earlier, the optimal speed for English speeches is 100-120 wpm, and one English word is equivalent to 1.3-1.7 Chinese characters (syllables). If we use the rate of 1.3:1, then 100-120 wpm will be equivalent to 130-156 Chinese characters per minute. If we use 1.7:1, then the equivalent Chinese speed will be 170-204 cpm. On the average, 150-180 cpm might be an appropriate rate. This is a much slower speed than the CCTV newscast.

5. Coping with Fast Speeches

Interpreters have to find coping strategies to deal with speeches that are delivered faster than the optimal speed. Based on the author’s practice and observation, four strategies may be used: the speaker is advised to slow down, the interpreter speeds up, summarisation, termination of service.

Strategy one: Request the speaker to slow down. This is the first possible tactic when a speaker is speaking too fast; however, one must remember that reminders seldom work, for speakers are either set in their speaking habits, or are always trying to cover too much within a limited time. After being reminded, a speaker will usually slow down for a sentence or two, before quickly forgetting the rules. Also, too-frequent reminders impede communication. A fast rate of speech is a universal challenge for interpreters. On one occasion, a NATO interpreter even tried seeking help on the AIIC website. He proposed installing a “word flow monitoring device” to monitor speakers’ speech rate (Communicate 1999). If a speaker speaks too fast, the device will automatically issue a warning to remind him/her to slow down. The author has his doubts about the device’s practicality. Some years ago, the interpreters’ control panel was installed with a ‘slow-down’ button, which interpreters could press when people were speaking too quickly. A small light affixed to the rostrum would blink, signalling to the speaker to slow down. The button’s disappearance from the control panel in recent years shows its ineffectiveness. Speakers often get carried away, and no amount of pressing on the button will alert them to their speed.

To ensure that speakers speak at a reasonable pace, the efforts of several players are indispensable: interpreters’ associations/societies, conference organisers, and individuals. As an association of interpreters, AIIC has communicated with institutional users of SI such as the United Nations and the European Union on speakers’ requirements. As a result of AIIC’s efforts, the European Commission’s Directorate-General for Interpretation issued Tips for Speakers, giving speakers tips on how to deliver speeches when interpreters’ services are used (Tips for Speakers). One of the tips is to “speak naturally, at a reasonable pace.” If speakers could observe the tips, interpretation quality would be greatly enhanced. In dealing with ad hoc users of interpretation services, interpreters should communicate with the conference organiser, requesting speakers to speak at a reasonable pace. At small and more personal meetings, interpreters may communicate directly with the speakers. Most speakers are quite ready to cooperate, and would even request interpreters to raise their hands at any time when slowing down is needed. As a general rule, interpreters must make known their difficulties to conference organisers and speakers to ensure that they receive the required cooperation.

Strategy Two: The interpreter speeds up. If the speaker fails to cooperate, the interpreter will have to speed up. The constraint, however, is that speech rate cannot be increased indefinitely. And as mentioned earlier, if a speech is delivered too quickly, most of the interpreter’s processing capacity will be devoted to comprehension, leaving little energy and time for translating and speaking the target language. The resulting utterance would be either incoherent or too fast for the audience’s comprehension.

Strategy Three: Summarisation. This is a tactic used when the interpreter cannot keep pace with the speaker even with increased speed. The best approach to interpreting speeches with redundant information (such as impromptu speeches) is to summarise. Summary interpretation will produce a more succinct speech than the original. On the other hand, speeches densely packed with information or with complex reasoning cannot be further condensed. Any attempt at summarising will result in omissions or truncated logic, and the audience will find difficulty following the speaker, particularly when speakers read verbatim and rapidly from written text or in government jargon. Interpreters can only do their best, and translate as much as they are able within the limited time. Having said that, summarisation consumes a lot of energy and is not sustainable.

Interpreters familiar with the subject-matter and who understand the audience’s needs and the central message can respond faster by capturing essential information and discarding the trivial. In any case, no amount of familiarity with the subject-matter can compensate for a speech delivered too fast.

Strategy Four: Termination of service. In the event that working conditions do not meet the interpreters’ minimum requirements and as a result, interpretation quality is compromised, interpreters may terminate their services. Although the author has not seen any interpreter using this strategy, the author has heard it. As a student at Beijing Foreign Studies University some years ago, the author was listening to a recording of debates at the United Nations General Assembly, and an interpreter was interpreting from another language into English. At the outset, the author had already found the interpreter’s delivery very fast. Halfway through interpretation, the interpreter burst out suddenly, saying something to the following effect: “Mr Chairman, interpreters are not machines. The speaker is speaking too fast. If you do not intervene, the interpreters will stop working.” Then it was complete silence. The interpreter had obviously turned off the microphone. Indeed, interpreters have rights, and are entitled to claim their rights. Unless the required working conditions are met, interpreters have the right to refuse to deliver service.

Having said that, interpreters who serve as the bridge between speakers and audience should refrain from turning off the microphone (despite it being one possible strategy) unless absolutely necessary. First of all, not all speakers speak fast, and any speaker’s turn will be over very soon. The organiser will understand if the interpretation for that particular segment is not completely satisfactory. Turning off the microphone because of one speaker will attract the audience’s attention to the interpreter. Secondly, many conferences provide SI services simply to raise the prestige of the meeting or as part of the complete language service offered at the conference. If the interpreter determines that this is indeed the case, i.e. meeting participants do not actually use the interpretation service, there is even less necessity to turn off the microphone and attract the organiser’s attention to interpreters. Finally, as a protection of the interpreter’s interests and to prevent any unnecessary dispute, interpreters should stipulate in their interpretation contracts that the organisers ensure that speakers observe a reasonable rate of speech.

  • Chang, Chia-Chien (2005), Directionality in Chinese/English Simultaneous Interpreting: Impact on Performance and Strategy Use. PhD thesis. Graduate School of the University of Texas at Austin. Online at (consulted 13.09.2006).
  • Communicate (1999). AIIIC Webzine n°1. Online at (consulted 18.10.2009)
  • Feng, Jingbao (冯京葆) (2002), 我做口译的经历和体会. (consulted 2006-9-13)
  • Gerver, David (1969). “The effects of source language presentation rate on the performance of simultaneous conference interpreters”. Emerson Foulke (Ed), Proceedings of the 2nd Louisville Conference on Rate and/or Frequency Controlled Speech. Louisville, KY: University of Louisville. October 22-24, 1969. 162-184.
  • Gile, Daniel (1995). Basic Concepts and Models for Interpreter and Translator Training. Amsterdam: John Benjamins.
  • Lederer, Marianne (1981). La Traduction simultanée. Paris: Minard Lettres Modernes.
  • Seleskovitch, Danica (1978). Interpreting for international conferences. Washington, D. C.: Pen&Booth.
  • Tips for Speakers (no date). Online at (consulted 18.10.2009)
  • VOA Special English (2005). Online at (consulted 18.10.2009).
  • Xie, Likui (谢礼逵) (2002),广播新闻播音语速浅析,中国主播网,Online at (consulted 13.09.2006)

Li C.Li Changshuan is Associate Professor, Graduate School of Translation and Interpretation, Beijing Foreign Studies University, China. He is also a conference interpreter and translator. Email:

Note 1:
Based on the author’s experience, with succinct writing, depending on the nature of the document, this ratio may be as high as 1.7:1.
Return to this point in the text