Survey on quality and role: conference interpreters’ expectations and self-perceptions

In the fall of 2008 we asked AIIC members to take part in a web-based survey on the two inter-related aspects of quality and the interpreter’s role in the communication process.

The Survey on Quality and Role, which is part of a larger research project on Quality in   Simultaneous Interpreting carried out at the Center for Translation Studies, University of Vienna, yielded 704 responses. This report summarises the main findings so as to share them with the AIIC community and express our gratitude to all members who kindly filled in the questionnaire.

Almost a quarter century ago, Hildegund Bühler (1986) first asked AIIC members, including seven members of CACL, what importance they attributed to various quality criteria when sponsoring candidates for AIIC membership, and her study inspired one of the most prolific and coherent lines of research on quality in interpreting research. Bühler’s data comprised barely four dozen responses but are nevertheless still cited to represent AIIC members’ viewpoint on quality. The quality criteria she listed in her survey became something like the backbone of empirical research on quality in conference interpreting. With the exception of Chiaro and Nocella (2004), however, follow-up surveys were targeted solely at users and yielded some rich findings1. No study on quality has since been addressed to the AIIC community, hence our survey project, which takes its inspiration from Bühler’s pioneering work.

In line with the growing interest in the sociological dimension of conference interpreting, our study also examines interpreters’ self-perception of their role in the communication process, as studied by Claudia Angelelli (2004) in a survey on role perceptions among conference, court and medical interpreters in North America. Angelelli set out to challenge the notion of the interpreter’s invisibility and to gauge the degree of an interpreter’s involvement in communication.

We see the question of the interpreter’s function and role as inherently linked with the issue of quality, since conference interpreters’ perceptions of the nature of their task will ultimately shape their performance.

Our main findings for the two interrelated issues of quality criteria and role perceptions are presented in the following synopsis. A comprehensive and detailed presentation and discussion of the survey results, complete with statistical analyses and figures as well as theoretical foundations, will be published in due course by the first author.

Methods

The Survey on Quality and Role, which is part of a larger research project on “Quality in Simultaneous Interpreting” at the Center for Translation Studies of the University of Vienna, was carried out as a web-based survey with a questionnaire generator tool. This approach is relatively new within the field of interpreting studies but clearly holds great potential (see Zwischenberger 2009). A pioneering example, which also served as the chief inspiration for the present study, is the survey on quality criteria by Delia Chiaro and Giuseppe Nocella (2004).

Survey design and questionnaire

The Survey on Quality and Role was conducted as a full-population survey among AIIC members in the fall of 2008. Before sending out the e-mail invitations, an extensive pre-test was conducted among three dozen conference interpreters who either work as freelance interpreters or permanent staff members for the EU institutions, but who are not affiliated with AIIC.

A total of 2523 e-mail invitations containing the survey link were sent out, of which 49 could not be successfully delivered. The list of e-mail addresses was compiled from the printed version of the 2008 AIIC Directory.

The survey was carried out with the help of LimeSurvey, a web-based questionnaire generator tool installed on one of the servers of the University of Vienna Center for Translation Studies. The software creates two separate databases – one containing all the e-mail addresses and the other containing all the responses. What is crucial is that the two databases are not linked. While the system allows the survey administrator to see who has submitted a response, the response as such cannot be related to the respondent.

Participants were informed of this safeguard of their anonymity before clicking on the “submit” button on the last page of the questionnaire.

For every entry in the e-mail database the system creates a personalized link to the questionnaire, ruling out multiple completions by the same or other persons.

The survey fielding time was exactly seven weeks (22 Sept. to 10 Nov. 2008), including two reminders. A total of 704 AIIC members completed the questionnaire, which corresponds to an impressively high response rate of 28.5%.

The questionnaire consisted of 41 items, including some follow-up questions, and was divided into three main parts. Part A elicited information on sociodemographic background variables, such as age, gender, working experience, and language combination, while Part B was essentially a replication of Bühler’s (1986) study on quality criteria. Part B also included a web-based experiment for which respondents were asked to listen to a one-minute audio sample of a simultaneous interpretation and give their impression. Part C was devoted to aspects of the conference interpreter’s role.

Research Questions

The survey focused exclusively on simultaneous conference interpreting. It sought to find out how AIIC members judge the importance of 11 output-related quality criteria (including fluency of delivery, correct grammar, lively intonation, logical cohesion, completeness, native accent, pleasant voice, and sense consistency with the original) and whether the relative importance of these criteria varied depending on the type of meeting or assignment.

Putting these ratings to the test, respondents were asked to listen to a one-minute sample of a professional simultaneous interpretation and rate it on a six-point scale. The survey recipients had been randomized into two groups, one hearing the sample with lively intonation and the other with monotonous intonation. The embedded experiment thus enabled us to check abstract ratings against actual quality judgements. As shown by previous research (e.g. Collados Aís 1998/2002), delivery-related quality criteria such as intonation are usually given less importance when rated generically, but play a significant role when the overall quality of a given interpreting performance is to be assessed. Our hypothesis was that even though the criterion “lively intonation” might receive relatively low ratings, the sample with lively intonation would be judged significantly better than the monotonous one, all other things being equal.

With regard to role, the survey examined how interpreters’ sociodemographic and professional background data related to their perception of four behavioral constructs (“Intervention in the original”, “Loyalty to the speaker/original”, “Reaction to working conditions” and “Interpreter’s detachment”) that had emerged from the analysis of role-related questionnaire items.

Results

Part A:  Sample / [Profile of] Respondents

Of the 704 conference interpreters who filled in and submitted the web-questionnaire, 76 percent were female, and 24 percent were male. Eighty-nine percent work as freelance interpreters, whereas only 11 percent of respondents are staff interpreters. With these ratios, the sample closely matches the membership structure of AIIC (see Neff 2008).

The average AIIC member in the sample is 52 years old, with a minimum of 30 and a maximum of 87 years of age. The largest group of respondents is between 50 and 59 years old. In terms of formal education received, 79 percent of respondents indicated having a university-level degree in interpreting/translation and 60 percent also hold a university-level degree in another field.

Participants’ average working experience as conference interpreters is 24 years, with a minimum of 4 and a maximum of 57 years. Most of the responses fell into the category of 20 to 29 years of working experience. As far as the working languages were concerned, the most frequently reported A language was French (24%), closely followed by English (22%) and German (18%). Not surprisingly, English (55%) is in the clear lead among B languages, followed by French (27%) and German (9%). The pattern is rather similar for C languages with English (47%) in the lead again, followed by French (43%) and then Spanish (29%).

Participants were also asked to indicate the sector in which they primarily work. The majority (42%) are primarily engaged in the non-agreement sector (private market), 33% work for the agreement sector (UN family, EU institutions, etc.), and the rest (25%) are evenly distributed between the two. When asked about the field(s) they primarily work in, 80.5% of respondents indicated the field of Business & Economics, closely followed by Politics (79%), Law (53%) and Technology (49%).

In terms of working mode, the vast majority of informants work primarily in the simultaneous mode: 79% of respondents indicated that they “rarely” or “never” work in the consecutive mode.

40.5% of respondents have been members of AIIC for up to nine years. The average is 15 years, with a minimum of one year and a maximum of 55 years of membership.

PART B: Quality Criteria & Audio Sample

Quality criteria

In Part B of the questionnaire, respondents were asked to rate the relative importance of 11 criteria for assessing a simultaneous interpreter’s performance on a four-point ordinal scale.

Table 1 summarizes the ratings. The criteria, most of which are Bühler’s, are presented in the same order as in the web-based questionnaire. Data from the present survey are shown in boldface, while Bühler’s (1986) results appear underneath in normal font.

Table1. Relative importance of output-related quality criteria (in percent).

 

very important

important

less important

unimportant

N=

Fluency of delivery

70.7
49

28.6
49

0.7
2

-
-

704
47

Correct terminology

61
49

38
51

0.9
-

0.1
-

703
47

Correct grammar

54.4
48

40.4
50

5.1
2

0.1
-

701
46

Sense consistency with original

88.3
96

11.1
4

0.6
-

-
-

702
47

Lively intonation

28.2

59.3

11.7

0.9

703

Native accent

14.1
23

42.1
47

39.7
28

4.1
2

701
47

Logical cohesion

74.8
83

24.8
15

0.4
2

-
-

698
47

Pleasant voice

27.5
28

58.5
61

12.7
9

1.3
2

702
46

Synchronicity

15.3

52

30.1

2.7

675

Appropriate style

36.2
17

55.6
68

7.4
15

0.9
-

702
47

Completeness

47.7
47

45.7
49

6.3
4

0.3
-

698
47

As in Bühler’s study, the criterion of sense consistency with the original received the highest ratings. It was considered very important by 88.3% of respondents, followed by the criterion of logical cohesion, which was rated as very important by almost 75% of survey participants. However, some of the respondents underlined in their comments that they themselves did not feel solely responsible for the logical cohesion of their interpretation: “Logical cohesion can only be a criterion when the original speech is coherent!” (R 23), “Logical cohesion depends on the speaker although we must do our best to improve its logic.” (R 129). Both criteria have lower percentages in the category “very important” when compared to Bühler (1986), whereas AIIC members responding to the 2008 survey seemed to be more demanding regarding the importance of such form-related criteria as correct terminology, correct grammar and appropriate style.

The parameter of fluency of delivery was attributed very high importance by almost 71% of survey participants and thus constitutes the leading criterion among all delivery-related criteria, followed by lively intonation (28.2% = very imp.) and pleasant voice (27.5%= very imp.). Synchronicity (15.3% = very imp.) and native accent (14.1% = very imp.) were rated the two least important quality criteria. Almost one third (32.8%) of participants rated synchronicity as either less important or unimportant. Quite a few respondents pointed out that the importance of the criterion of synchronicity varied with the type of discourse: “Synchronicity is important in certain types of speeches – with punch lines or lots of numerical data. In other speeches I would rate it as less important.” (R 643), “Synchronicity is important when finishing and at crucial moments such as when a joke is made but it should not take precedence over allowing the necessary distance to allow proper processing of the meaning.” (R 531) In the case of native accent as many as 43.8% of respondents considered the criterion either less important or unimportant. Some survey participants stressed in their comments that the importance of a native accent depended on the target language: “The native accent is more important when going into French than when going into English, where more flavors of English are customary” (R 550); “speakers of languages other than English are often intolerant of non-native accents” (R 338). In one very perceptive comment, accent was related to prosodic quality: “Native accent threw me, because if it was only accent it would be less important, but it is invariably associated with native intonation, which is essential to meaning” (R 137).

Mindful of the diversity of meetings at which conference interpreters work (e.g. Gile 1989), we also asked whether the importance of the quality criteria varied depending on the type of meeting or assignment (e.g. large assembly, training seminar, negotiation, press conference). 43.3% of respondents ticked the category “Yes”, 17.6% “Not sure, maybe”, while 39.1% opted for “No”. Those answering “Yes” or “Not sure, maybe” were requested to indicate spontaneously what might vary, when, and how. The three most frequently mentioned types of meetings with special quality requirements were the technical congress, media events and training seminars/workshops. In addition, respondents indicated that the importance of the previously rated criteria also varied with certain technical fields such as politics, business and economics or law etc. The degree of formality/informality, the duration as well as the size of a meeting were also mentioned as having an influence on the relative importance of the various quality criteria. According to these comments it is mainly the form- and delivery-related criteria that vary. Interestingly, it was the criteria which received a medium or lower overall rating, such as correct terminology, appropriate style or synchronicity, that were spontaneously mentioned as top priorities or of high importance when associated with concrete interpreting situations. Synchronicity, for example, was considered very important or even a top priority for media events by 46% of the 63 respondents who made reference to media settings in their comments. By the same token, 56% indicated appropriate style as a top priority or of high importance for political settings, and correct terminology was considered a top priority for seminars/workshops by 38% of those 45 respondents who cited this meeting type. The two criteria with the highest overall ratings, sense consistency with the original and logical cohesion, were generally deemed to be equally important for all interpreting situations: “all of [the criteria] vary, except sense consistency and logical cohesion” (R 430).

Audio sample

The sample of a simultaneous interpretation into English, by a young professional interpreter with German as her A language, was assessed on a six-point numerical scale (6=very good; 1=very bad) by a total of 567 survey participants. Some of those who could not or did not give an opinion on the sample reported that they were either unable to listen to the embedded audio file for technical reasons or considered it unfair or impossible to judge a colleague’s performance without knowing the original: “Really unfair to judge her, it all depends on the original, working conditions, sound, etc.” (R 168); “I really can’t judge the interpretation without the original” (R 207). The results for the 567 respondents who did rate the sample are shown in Table 2.

Table 2. Comparative ratings of audio samples

Sample with

N

Mean

Standard
deviation

Lively intonation

290

3.84

0.99

Monotonous intonation

277

3.68

1.06

Total

567

3.76

1.03

The difference in mean ratings between the two audio samples is small (0.16) but nevertheless statistically significant (t-test). This confirms once again that the criterion of intonation is by no means negligible and has an appreciable impact on judgements of overall interpreting quality, even when listening for only one minute. Presumably, the impact of monotonous intonation would be even more pronounced for a standard turn length of up to 30 minutes.

Further statistical analyses showed that gender had a statistically significant effect on the way the audio samples were assessed. Women gave a more favourable total mean rating than men (female interpreters: 3.82 vs. male interpreters: 3.58), and the difference between men and women in their rating of the lively and the monotonous sample was significant (lively intonation: female interpreters: 3.92 vs. male interpreters: 3.57; monotonous intonation: female interpreter: 3.71 vs. male interpreters: 3.60). Interestingly, male interpreters rated the interpretation with monotonous intonation (3.60) slightly better than the interpretation with lively intonation (3.57). On the whole, these findings suggest that women are more generous judges than men and that female interpreters are also more receptive to the presence or absence of the parameter of lively intonation.

Respondents’ A language also had a statistically significant impact on the rating of the two audio samples. A comparison between the three most frequently reported A languages (French, English and German) showed that AIIC members with English A gave the strictest total mean rating in both conditions and differed significantly from the ratings given by participants with French or German as their A language (English A: 3.30, French A: 3.67 and German A: 3.84). Respondents with German A gave the most favourable rating in both experimental conditions.

Further significant effects were found for the two background variables of age and working experience, for which respondents were divided into three groups. The youngest third of AIIC members in the sample (30-47 years) rated the sample with monotonous intonation significantly better (3.84) than the middle (48-57 years: 3.53) and oldest thirds (58+ years: 3.63). There was a higher degree of consensus for the audio sample with lively intonation, for which the youngest and the oldest third gave almost identical ratings (3.86 vs. 3.84), while the middle third gave it a somewhat lower rating (3.74). This seems to indicate that younger interpreters are not as receptive to the presence or absence of the parameter of lively intonation as older professionals.

The same applies to participants’ working experience. Again the group with shorter working experience (4-19 years) gave the most generous rating (3.84) for the interpretation with monotonous intonation and differed significantly from the other two groups (20-29 years: 3.59; 30+ years: 3.60). The ratings of the sample with lively intonation show a different pattern. Here it is the third with the most working experience (30+ years) that gave the best rating (3.91) compared to the other two groups.

A further significant effect was found for the background variable of university-level degree in interpreting/translation. Participants with such a degree rated the sample with lively intonation significantly better (3.91) than those without (3.61), whereas the sample with monotonous intonation was rated almost identically by the two groups. Thus, formal university training in the field of interpreting and translation seems to be associated with more appreciation for lively intonation in an interpreter’s performance.

PART C: Conference interpreter’s roles

Conference interpreters’ self-representation

In an introductory question, interpreters were asked to describe their role as a conference interpreter in a word or two, or in a phrase. A total of 628 survey participants (89.2%) gave a spontaneous answer to the question.

The majority of respondents saw themselves in an assisting or helping function, using descriptions such as facilitator/enabler of communication (21.2%) or aid/helper for communication (8.9%), followed by those who described themselves as mediator/intermediary (13.7%) or bridge/link (10.7%). Others saw themselves as professionals rendering a communication service, using descriptions such as communicator (11.5%), or expert/(communication-)service provider (2.1%) who makes communication work (5.6%).

The descriptions and definitions elicited in this study reflect the self-representations produced by influential pioneers and AIIC itself and can be regarded as socio-professional norms which reflect the values of the community of AIIC conference interpreters2.

Role components

In Part C of the questionnaire, participants were asked to express their agreement or disagreement with 14 role-related statements on a six-point numerical scale (6 = completely agree; 1= completely disagree). In a preliminary analysis, these 14 items were found to correspond to four distinct role components, labelled as “intervention in the original speech”, “loyalty to the speaker/original”, “reaction to working conditions” and “interpreter’s detachment”. The aim was to establish whether there was a correlation between the sociodemographic background of respondents and their perception of the four role components.

For the first role component, “intervention in the original speech”, statistically significant differences were found for the background variables of age and working experience. The three age and working-experience groups created for the purpose of the analysis differed significantly from each other. The items underlying Component 1 were consistently given the highest agreement by the oldest third of respondents (58+ years) and the group with the longest working experience (30+ years); conversely, the lowest agreement was found among the youngest third of respondents (30-47 years) and the group with shorter working experience (4-19 years). Thus, AIIC members’ stated readiness to intervene in the original speech by moderating a speaker’s words when they clash with cultural conventions, using their own language and style when interpreting, trying to ensure that the interpretation is intelligible even if the original is not, etc. increases proportionally with age and working experience.

Role component 2, “loyalty to the speaker/original”, showed significant differences with regard to gender. Female interpreters expressed a higher level of agreement with the need to ensure that their interpretation reflects the speaker’s tone and register as closely as possible and that their interpretation has the same effect as the original etc.

This finding is in line with another result of the study. When respondents were asked, in an earlier item of Part C, to rate the importance of feedback from speakers and other reference groups, female interpreters attributed a significantly higher degree of importance to the speaker’s feedback than male professionals did: feedback from speakers was rated as (very) important by 72.1% of women respondents, compared to 59.9% of men. Thus, loyalty (fidelity) to the speaker and the original appears to be a hallmark of female professionals more so than of their male colleagues.

No background-related differences were found for Component 3, “reaction to working conditions”, which suggests that all interpreters, irrespective of age, gender or working experience, invariably find it essential that their needs (for documents, breaks etc.) are respected. In contrast, age affected responses regarding Component 4, “interpreter’s detachment”, which comprised items such as “My professional distance as an interpreter keeps me from being influenced by emotional events in the meeting room”. The youngest third of respondents (30-47 years) disagreed significantly more with the notion of the interpreter’s detachment from the communication process than the other age groups.

On the whole, our survey shows AIIC members to be a rather homogeneous group with regard to their attitude toward the various role-related questionnaire items and the higher-order constructs derived from them, that is, intervention, loyalty, reaction to working conditions, and detachment.

Conclusion

The Survey on Quality and Role largely confirms Bühler’s (1986) findings for the relative importance of various quality criteria, especially with respect to the two content-related criteria sense consistency with the original and logical cohesion. While these were rated not quite as highly as in Bühler’s study, participants in the Survey on Quality and Role were more demanding for the form-related quality parameters of correct terminology, correct grammar and appropriate style. Delivery-related criteria, in contrast, were attributed a lower degree of importance – except for the criterion of fluency. However, when respondents were asked to relate the criteria to particular meeting types or domains, many of the parameters that received a medium to low overall rating were spontaneously mentioned as very important or even as top priorities. This highlights that the degree of importance attributed to various quality criteria varies in relation to the meeting (type of event, degree of formality, duration and size) and the domain (political, diplomatic, business, legal, medical) in which the event takes place.

As revealed by the experimental audio sample embedded in the web-based questionnaire, professional interpreters’ assessment of a simultaneous interpretation is affected to a significant degree by the isolated variable of intonation, even though “lively intonation” is not among the seven criteria that are rated as “very important” by at least a third of the respondents. Moreover, interpreters’ impression of a fellow practitioner’s output is shaped by sociodemographic background variables such as gender, age and working experience. Thus, female interpreters are more appreciative of lively intonation and are also more generous judges than male professionals, whereas  younger and less experienced professionals are less receptive to the absence or presence of lively intonation.

On the assumption that interpreters’ views on what makes for a good simultaneous interpretation are shaped by the way they perceive their role in the communicative event, survey participants were asked to give a spontaneous description of their task and to express their agreement or disagreement with fourteen role-related statements underlying the constructs of “intervention”, “loyalty”, “reaction to working conditions” and “detachment”. Both in their self-descriptions – in such terms as “facilitator/enabler of communication”, “mediator/intermediary” or “bridge/link” – and in their behavioral attitudes, AIIC members are shown to share a clearly defined professional ethos. This suggests an effective process of professional socialization, presumably driven by university-level training, interaction with peers, and not least by a highly active professional organisation. Nevertheless, the survey reveals some significant relationships with sociodemographic background variables. Respondents’ stated readiness to “intervene” in the original, for instance,  increases with age and working experience, and women show a higher degree of loyalty to the speaker/original than their male colleagues.

These findings from the Survey on Quality and Role, provided here in summary form, consolidate and enhance our understanding of conference interpreting as a profession, thanks to the readiness of AIIC members to share their experience and expertise. We would therefore like to express our sincere gratitude to all those who kindly filled in the questionnaire, and would be happy if they, too, felt that the time and effort spent on this survey, on their as well as on our part, have proved well justified by the insights gained.

Footonotes

  • A detailed summary of those results as well as a discussion on quality research in general can be found in Kahane (2000).

  • For a detailed analysis, see Zwischenberger (2009).

References

Angelelli, Claudia V. (2004) Revisiting the interpreter’s role. A study of conference, court and medical interpreters in Canada, Mexico, and the United States. Amsterdam/Philadelphia: John Benjamins.

Bühler, Hildegund (1986) Linguistic (semantic) and extralinguistic (pragmatic) criteria for the evaluation of conference interpretation and interpreters. Multilingua 5 (4), 231-235.

Chiaro, Delia & Nocella, Giuseppe (2004) Interpreters’ perception of linguistic and non-linguistic factors affecting quality: A survey through the World Wide Web. Meta 49 (2), 279-293.

Collados Aís, Angela (1998/2002) Quality Assessment in Simultaneous Interpreting : The importance of nonverbal communication. In: F. Pöchhacker/M. Shlesinger (eds.) The Interpreting Studies Reader. London/New York: Routledge, 326-336.

Gile, Daniel (1989) Le flux d’information dans les réunions interlinguistiques et l’interprétation de conférence: premières observations. Meta 34 (4), 649-660.

Kahane, Eduardo (2000) Thoughts on the quality of interpretation. http://www.aiic.net/ViewPage.cfm/page197 (accessed 20 January 2010).

Neff, Jacquy (2008) A Statistical Portrait of AIIC: 2005-06. http://www.aiic.net/ViewPage.cfm/article2127.htm (accessed 18 January 2010).

Zwischenberger, Cornelia (2009) Conference interpreters and their self-representation: A worldwide web-based survey. Translation and Interpreting Studies 4 (2), 239-253.



Recommended citation format:
Franz PÖCHHACKER,Cornelia ZWISCHENBERGER. "Survey on quality and role: conference interpreters’ expectations and self-perceptions". aiic.net March 15, 2010. Accessed November 14, 2018. <http://aiic.net/p/3405>.



There are no comments to display