How interpreting is perceived by users

AIIC's Research Committee organised a survey of how the interpretation service is perceived by users. Jennifer Mackintosh, Project Co-ordinator, reports.

The Users Expectation Survey was co-ordinated by the AIIC Research Committee (RC). On the basis of an initial brief from the RC, Peter Moser of SRZ Stadt- und Regionalforschung of Vienna designed and tested the questionnaire and analyzed the results, which reflect the views of over 200 participants at 84 conferences in 25 countries.

According to the replies obtained in the course of quite lengthy interviews (15-20 minutes), the ‘ideal interpreter speaks in a clear and lively manner but doesn’t go in for histrionics, understands the subject matter of the meeting and knows the terminology, speaks in complete, grammatically correct sentences and enunciates clearly without ‘umming’ and ‘aahing’. Clarity of expression is the quality users rate above all others. The ‘ideal’ interpreter puts faithfulness to the meaning of the original speech above other considerations and in so doing concentrates on essentials rather than trying to reproduce literally everything that is being said. Delivery is regular and, surprisingly high on the list of points raised spontaneously by respondents, the ‘ideal’ interpreter keeps as close as possible to the original. 34% of the interviewees said that they are uncomfortable if the interpreter waits for a long time before starting to interpret, or pauses for long periods during a speech. Non-standard accents (regional or non-native) do not appear to affect users very much. This could be because at meetings using AIIC interpreters listeners would virtually never be exposed to non-native or very marked regional accents. It might also be because if, exceptionally, they hear a strong foreign accent they may well be more bothered by incorrect syntax or grammar than by the accent. Whatever the reason, when asked about accents 50% of the respondents rated them unimportant and a further 24% said they were not really irritating. Only 8% said they were very irritating and 14.4% found them irritating (the rest were ‘don’t knows’). Looked at by respondents’ language group the picture is (in % of responses for that language group):

  very irritating fairly irritating not really irritating unimportant
English 9 14 19 58
French 2 10 29 58
German 0 17 33 50

The higher sensitivity to accents in the English-speaking group may be explained by the fact that they are the most frequently exposed to a wide range of non-native and regional accents, especially from other delegates. The numbers of respondents for the other language groups was not large enough for the replies to be representative.

The questionnaire used by the interviewers was quite detailed and followed a rigid pattern. For reasons of cost the interviews were conducted by interpreters, not professional interviewers which meant that very little scope to vary the interview could be allowed.

There were some partly open-ended questions such as Question 2 which asked interviewees to enumerate requirements and expectations of interpretation over and above the three factors expressly mentioned (completeness of rendition, clarity of expression, and correct terminology). They were asked to rate these either as ‘very important’, ‘fairly important’, ‘fairly unimportant’, ‘unimportant’ or ‘don’t know’. The three were rated ‘very + fairly important’ as follows:

  • clarity of expression: 97.3
  • correct terminology: 87%
  • completeness of rendition: 86.6%

The requirement most frequently raised spontaneously was synchronicity (i.e. avoiding too great a time lag between the original and the interpretation), occurring 25 times (i.e. in 12.4% of the responses). All in all 91 respondents made 141 spontaneous references to aspects of content (such as understanding of subject, accuracy, faithfulness to the original, terminology), and over the interviews as a whole 68 respondents made 89 references to synchronicity. The same number of respondents made 85 references to rhetorical skills (for example, regular delivery, absence of hesitation, complete grammatically correct sentences, clarity of expression) and the same number again made 93 references to voice features (lively, non-monotonous, naturalness, reflecting the feeling of the original). 8 respondents were critical of an over-emphatic or flamboyant interpreting style and there were a total of 21 references to thedesirability of adopting a neutral tone in relation to what is being said.

A number of interviewees (31) raised poor mike discipline as a concern. They said they were disturbed by failures to use the cough button, the sound of papers being shuffled, tapping on the mike or table top, whispering or laughter in the background. Almost as many (26) mentioned the problem of technical failures. 10 expressed the original, if slightly impractical, idea that in the event of an equipment breakdown the interpreter should briefly summarise what had been said while the system was off the air as soon as it came back on... which does rather assume that the interpreters themselves had not been affected by the breakdown.

When respondents were asked to state which feature was the most important to them: faithfulness to meaning of the original or a literal reproduction of what was said, 66.2% replied faithfulness to meaning, 14.9% opted for a literal reproduction, 9% said both were equally important and for 6.5% it depended upon circumstances. A similar question asked interviewees to rank the comparative importance of focusing on essentials vs. giving a complete rendition (question suggested by data obtained by Ingrid Kurz, cf. 1993, Conference Interpretation: Expectations of Different User Groups, The Interpreters Newsletter No 5, pp. 13-21, SSLM, University of Trieste), 43.8% replied that the interpreter should focus on essentials and 31.8% opted for completeness. In themselves these results are already quite useful in confirming some of our assumptions (e.g. regarding faithfulness to meaning) but after detailed analysis by SRZ on the basis of variables such as age of respondent, conference-going experience, type of conference attended, they yield quite a lot more information. There is, for example, a big difference according to the age of the respondents: some 74% of the over-60 year olds put faithfulness to meaning first whereas those under 30 rank it equally with literal reproduction. However this group was very small and contained the least experienced conference-goers of all the age groups (under 30, 30-45, 46-60, 60+).

The basic assumption the study set out to explore, informed by earlier studies conducted by I. Kurz and others, was that different user groups would have different requirements. The 84 conferences (giving, with sub-sessions, a total of 100 individual meetings) at which interviews were conducted were divided up into four categories: small technical (44%), large technical (29%), small general (12%) and large general (15%). Although some of the differences between the groups are unsurprising, e.g. the importance attached to terminological accuracy is highest in technical meetings, there are others which are less predictable. For example, 80% of the respondents at small general meetings ranked faithfulness to meaning above literal reproduction whereas just under 60% of the participants at large general meetings did so. Is it because the latter are often highly ritualized annual general conferences where the medium is an important part of the message? Whereas the former are frequently business meetings involving a number of decisions?

Differences according to conference experience and meeting type emerge for the importance of terminological accuracy. As mentioned above, participants at technical meetings attach greater importance to terminology than those at general meetings. This is particularly evident if the number rating it ‘unimportant’ are looked at: 17% of the respondents at general meetings said that terminological accuracy was unimportant against only 7% at technical meetings. Also the older the respondent, the greater the likelihood of terminological accuracy being rated very important.

In order to further test the hypothesis that different user groups would have different expectations, respondents with earlier experience of conferences of a different type were asked about their expectations at those meetings. 60 of the respondents fell into this category and the distribution was virtually the same as in the main group. Expectations were virtually the same: 90% maintained their preference for faithfulness to meaning over literal reproduction but 13% of those preferring concentration on essentials at the current meeting opted for completeness when asked about the earlier conferences.

The questionnaires also sought to explore attitudes towards using interpretation, towards interpreters as well as perceptions of how interpreters work. When asked about their headsets 77% of the respondents said they were quite satisfied but 21% were not. 55% said that they had already taken a look inside an interpreter’s booth. When asked how long they thought an interpreter should work without a break, 22% said up to 20 minutes, 28.5% up to 30 minutes but 6% replied for up to an hour and 5.5% for the length of a presentation, suggesting that changes of voice and style during a speech can be disturbing to some users.

Respondents were then asked what profession comes to mind if asked to compare the demands it makes upon its practitioners with those made on interpreters. The replies varied widely but show that interpretation is seen as a fairly stressful and demanding occupation. For example, pilots/air traffic controllers were mentioned 16 times; surgeons/emergency department medics 6 times; racing car drivers twice; also reporter/journalists 21 times, translators 18, actors 8, teacher/educators 7. When asked what they considered to be particularly difficult about the profession, 60 references were made to concentration and stress, 40 to the need for interpreters constantly to update their knowledge, speed was mentioned 28 times and 32 references were made to simultaneous listening, processing and speaking. It would appear that users of interpretation have a quite high level of awareness about what is required of an interpreter.

A set of questions on language and interpretation use, mother tongue and nationality showed that 117 of the 201 respondents (58%) did not require interpretation if English was spoken with 62 of the respondents giving English as their mother tongue (38 UK nationals, 14 US, 8 Canadians - not necessarily English mother tongue - and 1 South African). If French was spoken 92 participants required no interpretation, of which 50 gave French as their mother tongue (33 French nationals). 44 required no interpretation from German (30 German mother tongue), 26 needed none for Spanish (18 mother tongue) and apparently only Italian nationals (7) were able to follow in the original. There were 4 Swedish nationals in the respondent group but 10 could follow in Swedish (there was 1 Finn). Altogether 40 different nationalities were indicated and 13 languages specifically mentioned as mother tongues with 12 respondents not specifying. 65 participants said they listen, if only occasionally, to interpretation from English, 50 from French, 32 from German, 26 from Italian, 23 from Spanish. Nearly 50% of them sometimes listen to the interpretation even if they have fully understood the speaker and 7.5% said they nearly always do. If they pick up a mistake 24.5% doubt their own linguistic abilities, 20% doubt the interpreter’s reliability and 25.5% doubt both. When delegates are asked whether interpreters are sometimes blamed for the poor quality of a speech they are divided: 31.8% said definitely not and 17.9% said very often.

At the end of the Questionnaire interviewers were requested to provide data about the facilities (equal split between portable and permanent booths), equipment (65% of the systems were radio or infra-red), documentation (in 69% of the conferences documentation was sent to the interpreters before the meeting and in 77% material was distributed during the conference). Documents were predominantly available in English (75% of the meetings), French (55%), German (28.5) and Spanish (15%). At 3 meetings were no documents available. In only 5 cases were the interpreters given a briefing.

A total of 91 colleagues conducted interviews, following detailed interview notes prepared by P. Moser of SRZ. Their efforts will serve the profession as a whole as well as providing interpreter trainers, recruiters in their dealings with clients, chief interpreters and our Agreement Sector negotiating colleagues with valuable data . Our thanks are addressed to them and to the colleagues who translated the Questionnaire (from German to English, French, Italian and Spanish) and the Briefing Notes. Our particular thanks however go to Peter Moser who was unstinting in his efforts on our behalf and whose skill and experience have served us so well. The enthusiastic and appreciative reception the report has received are in themselves evidence of its relevance and value. The results enable us to state with some degree of certainty what users of simultaneous interpretation look for in the ‘ideal’ interpreter and which are the pitfalls to be avoided.

Recommended citation format:
Jennifer MACKINTOSH. "How interpreting is perceived by users". March 11, 2004. Accessed May 22, 2019. <>.

There are no comments to display