Vol. VI, No. 1 - Winter 1997

New Study on Fatigue Confirms
Need for Working in Teams

The practice of having simultaneous interpreters work in teams of two during lengthy assignments, although standard procedure in all other forums requiring interpretation, has never been universally accepted by the courts. In most state and many federal courts, it is simply not done. Attempts by interpreters to institute the policy have met with resistance from judges who consider it wasteful and administrators who cite budgetary constraints. But a study recently conducted at the University of Geneva has contributed important new information on the subject: its findings provide further scientific evidence to support the position that accuracy is directly related to the length of time that a person interprets.

The study by Barbara Moser-Mercer and her colleagues (forthcoming) at the University of Geneva’s École de Traduction et d’Interprétation constitutes the first part of a two-part study on stress and fatigue in conference interpreting. Its aim is to examine the fatigue factor during extended turns, as well as the coping behavior of interpreters when under stress. The subjects—five native English-speakers working from German into English, whose professional experience ranged from 12 to 25 years in the booth—were told to work until they could no longer provide acceptable quality. During the first 30 minutes the frequency of errors—as measured with an elaborate error scale—rose steadily. The interpreters, however, “appeared to be unaware of this decline in quality,” according to the report, as most of them continued on task for another 30 minutes.

The error scale included several different categories by which quality can be determined. “Looking at the total number of errors,” the report states, “we can see that the frequency increases from three minutes to 30 minutes.” The category of most serious errors, i.e., errors in meaning, rose consistently with increased time on task. At 60 minutes, all subjects combined committed a total of 32.5 meaning errors. “Considering that each meaning error, no matter how minor, does distort the message, a considerable increase in the number of meaning errors after 30 minutes on task does represent a significant decline in output quality,” the authors argue. In the category of nonsense, the number of errors committed by the subjects almost doubled after 30 minutes on task—from 4.5 after 15 minutes to 8.5.

Moser-Mercer and her colleagues conclude:

The increase in the number of meaning errors combined with the interpreters’ lack of awareness of this drastic decrease in quality shed some light on the validity of interpreters’ judgement of their own output quality [...] This lack of judgement appears to be the result of cognitive overload: a situation in which the interpreter tries to economize on processing capacity and allocate resources only to those parts of the interpreting process that will ensure continuous output (irrespective of the quality provided) [...] We can conclude from this that shorter turns do indeed preserve a high level of quality, but that interpreters cannot necessarily be trusted to make the right decision with regard to optimum time on task.

This is an important insight, since many interpreters, fearful of not getting work or of exposing what is erroneously perceived as a weakness, will insist that they can work for extended periods of time without any adverse consequences to accuracy. It also shows that some courts beg the question: if interpreters themselves are unable to judge the length of time beyond which the quality of their performance declines significantly, how can anyone else have the power to decide how long an interpreter should work without relief?

An additional conclusion reached by the University of Geneva team concerned the subjects’ emotional response to increased time on task. “Interpreters seem to experience an increase in stress during the first 30 minutes, as indicated by a rise in cortisole levels, but with task overload respond with an 'I couldn’t care less' feeling,” they report, adding: “This is borne out by anecdotal evidence according to which interpreters try to deflect responsibility for the quality of output when they consider the demands to be unrealistic; this would include increased time on task, extremely fast speakers, and long working hours.” Every court interpreter, no matter how experienced, would undoubtedly corroborate this finding.

Stress investigated among UN interpreters

H. McIlvaine Parsons, a fellow at the Institute for Behavioral Research, in Silver Spring, MD, reached similar conclusions in a consultation he conducted in 1975 for the United Nations. The study was part of an investigation that followed a job action in which UN interpreters stayed away from their jobs for one day to protest “working hours and the stress and tension they said resulted from working more than seven half-day sessions per week.” McIlvaine Parson’s objective was in part to “create a wider understanding than there seemed to be of the interpretation process. If some of these factors could be ameliorated,” he argues, “the interpreters might experience less stress and tension and they might be less likely to avoid that stress and tension by failing to come to work.”

McIlvaine Parsons reported that “the interpreters were emphatic that more than three hours in a booth [taking turns with a colleague] resulted in excessive stress and tension, especially compared with a shorter time.” Other factors rated by the subjects as stressful or extremely stressful included: the speaker talking very fast, lack of clarity or coherence by the speaker, the need for intense concentration, inexperience with the subject matter, a speaker’s accent, long speaker utterances between pauses, background noise in the meeting room, and mispositioning of the speaker’s microphone relative to the speaker. All of these would be equally applicable to court interpreters.

As a result of his study, McIlvaine Parsons recommended to the UN Secretariat “that a simultaneous interpreter should not be required to work more than three half-day sessions in succession.” It should be borne in mind that UN interpreters work in teams of two at all times. Skeptics might be inclined to argue that these studies do not refer specifically to interpreters who work in court and are therefore not applicable to this sector. A comparison of court and conference interpreting, however, can easily demonstrate that the former is in fact more demanding and stressful than the latter.

What is fatigue?

Although the definition of the word fatigue seems obvious, there is considerable confusion among the general public and the legal profession about its meaning and consequences in a courtroom setting. Fatigue for interpreters is not primarily physical, as in the case of athletes, whose muscles become strained after sustained exertion: it is mental fatigue. It results from complex mental processing and the high degree of concentration the interpreter must have to hear, then understand, analyze and finally express ideas coherently in another language. “Most people do not realize that an interpreter uses at least 22 cognitive skills when interpreting,” states Patricia Michelsen in an article published in The Court Management and Administration Report. Other studies of simultaneous interpretation have shown that fatigue is exacerbated by environmental factors that interfere with various aspects of the cognitive process.

Taking into consideration both cognitive processes and environmental interference, the degree of concentration required of an interpreter is many times greater than that of any other person in a courtroom. In a 1995 study on fidelity assessment in consecutive interpretation, Daniel Gile reports that a group of subjects asked to rate an interpretation “were found to be unreliable fidelity assessors: they did not detect all interpretation errors on the one hand, and imagined errors that had not been made by the interpreter on the other.” This is not surprising to interpretation teachers, according to Gile, since “ordinary listening entails too much loss, and [...] interpreters have to listen to speakers with much more concentration than is usual in everyday life.”

While conference interpreters must cope with the stress generated by the job’s cognitive demands, their booth-enclosed environment is relatively stress-free compared to a courtroom setting. As Michelsen indicates, “Conference interpreters work under better conditions: they concentrate on only one speaker at a time, often have a prepared text of the speech ahead of time, address the audience in only one level of rhetoric, and usually do not have audibility problems.”

Environmental factors and loss of accuracy

Audibility is one of the key factors contributing to the stress suffered by court interpreters. In 1974, an enlightening study on the effects of noise on the performance of simultaneous interpreters was conducted by David Gerver, then at the University of Durham, Great Britain. He found that, as the listening conditions deteriorated, significantly more errors where committed by the subjects when interpreting than when shadowing (repeating a spoken text in the same language).

This finding, according to Gerver, “suggested that difficulty in perceiving source language passages reduced the ability of simultaneous interpreters to monitor their own interpretations into the target language.” He added that other studies indicated that “levels of noise which would not necessarily impair perception of speech by simultaneous conference interpreters could interfere with the processes involved in the retrieval and transformation of the messages being interpreted.” Listening conditions are most relevant to any discussion of interpreter stress and fatigue. Since monitoring their own utterances and making corrections is one of the many cognitive functions performed by interpreters, if their ability to self-correct is impaired, their level of stress and resulting fatigue also increase proportionately. “It is perhaps not surprising,” Gerver comments, “that simultaneous interpreters are particularly sensitive to environmental noise and that they will often refuse to work in conditions which, to the observer at least, do not appear particularly stressful.”

While Gerver’s study was conducted with a monitored increase in noise level, the same conclusions would apply to a situation in which the interpreter is simply unable to hear, as too often occurs in the courtroom. Given that acoustic impairments cause conference interpreters stress and fatigue, we can safely conclude that court interpreters are at a distinctly greater disadvantage acoustically, and therefore subjected to even more severe stress. Unlike conference interpreters, who work in soundproof booths and hear the sound through headphones connected to a stationary microphone, court interpreters hear telegraphic, often-interrupted messages from speakers distributed throughout the courtroom. Although many courts have microphones, they are not multi-directional and often distort the sound more than they amplify it. The interpreter must then filter this message through myriad other noises polluting the audible space, such as telephones ringing, jurors coughing, babies crying in the gallery, and so on. The best kept secret in the courtroom may well be that interpreters are often unable to hear what they are expected to interpret. When interpreting simultaneously into a microphone, they are invariably made to position themselves at the point furthest away from the witness stand, so as not to disturb jurors and those testifying. When no simultaneous equipment is available, the interpreter is obliged to sit next to the defendant—the hardest place from which to hear the proceedings. (By contrast, court reporters are granted the choice spot in the well of the courtroom to maximize their ability to hear every word uttered.) Moreover, no one seems to realize that the interpreter’s hearing is further obstructed by the sound of his or her own voice overlapping the original speaker’s at all times, creating an additional acoustical impediment. The bolder or more experienced interpreters will interrupt to insist that the parties speak up or rearrange themselves to improve audibility. But courtroom atmospheres are not always conducive to intransigence on the part of someone who is supposed to be invisible and unobtrusive, and even well-meaning judges and court clerks often have little or no control over antiquated sound systems or acoustically faulty architecture.

All of the factors found by the various studies described here to be major causes of conference interpreter stress and fatigue—acoustics, prolonged periods on task, lack of familiarity with relevant terminology, excessively fast or incoherent speakers, etc.—are in fact more applicable to interpreters in court than in any other setting. Moreover, judiciary interpreters have the additional pressure of knowing that nothing less than the life and liberty of human beings are at stake in the proceedings they are called upon to duplicate in a defendant’s native tongue. The awareness that each word mistranslated or omitted hinders the non-English speakers’ ability to follow the proceedings against them is a constant source of tension. Whereas the conference setting allows for much more flexibility, interpreting in court requires greater precision, since a complete and faithful rendition must include hesitations, false starts, repetitions and inaccuracies. It follows then that judiciary interpreters face more demanding and stressful working conditions than their counterparts elsewhere.

Studies corroborate empirical evidence

The only way to ensure a faithful rendition is to provide interpreters with relief at half-hour intervals

While these studies make an important contribution to the body of scientific data needed for a better understanding of the interpreting process and its complexities, they merely corroborate what practicing interpreters have known and argued all along: that work quality—i.e., accuracy and coherence—begins to deteriorate after approximately 30 minutes of sustained simultaneous interpreting, and that the only way to ensure a faithful rendition of legal proceedings is to provide interpreters with adequate relief at approximately half-hour intervals.

Conscientious administrators in several federal courts, the United Nations and the U.S. State Department recognized the need for tandem interpreting and adopted the practice early on. Team interpreting, in fact, dates back to the Nuremberg trials. At the State Department, which according to Harry Obst, Director of the Office of Language Services, handles 200 to 300 interpreting missions in 100 different locations per day, it is considered an inviolable policy. In response to a request from Ed Baca of the Administrative Office of the U.S. Courts, Obst pointed out that “The policy on simultaneous interpreters is simple and corresponds to that of all other responsible interpreting services in the entire world (United Nations, European Commission, International Red Cross, International Court of Justice, foreign ministries in other nations.) No individual simultaneous interpreter is allowed to work for more than 30 minutes at a time.” The letter continues, “This is also done for the protection of the users. After 30 minutes the accuracy and completeness of simultaneous interpreters decrease precipitously, falling off by about 10% every 5 minutes after holding a satisfactory plateau for half an hour.” The reason, Obst explains, is that “The human mind cannot hold the needed level of focused concentration any longer than that. This fact has been demonstrated in millions of hours of simultaneous interpretation around the world since 1948. It is not a question of opinion. It is simply the result of empirical observation.”

Echoing the results of the University of Geneva study, Obst adds that although some interpreters believe they can interpret longer than that, they do so because after 30 minutes “they can no longer differentiate between interpreting the original message or just babbling in the target language. Their mind is too tired to evaluate their own performance.” The policy on the part of court administrators that interpreters work for an hour or more without relief, says Obst, “makes sense only in budgetary terms. It makes reliable interpreting impossible and denies the client who has to rely on the interpreter the due process that every person is entitled to under our laws.”

And that is precisely the point. Unlike their colleagues in any other sector, judiciary interpreters are placed under oath to “truly and accurately interpret” the proceedings. Accuracy in a legal context is not an academic concept or an abstraction that can be quantified in relative terms. It is the cornerstone that guarantees limited-English litigants equality under the law. That was the spirit of the Court Interpreters Act enacted in 1978. It is also the spirit of the Code of Professional Responsibility drafted by the Administrative Office of the U.S. Courts, which compels interpreters to “fulfill a special duty to interpret accurately and faithfully” and “perform to the best of their ability to assure due process for the parties” and “refuse any assignment [...] under conditions which substantially impair their effectiveness.” If interpreters are to be expected to comply with these canons, they will need the full support of administrators in both the state and federal courts, who will place due process considerations above the temptation to trim their budgets at the expense of those who come before the bar of justice.

References

Dueñas González, Roseann, Victoria F. Vázquez and Holly Mikkelson. 1991. Appendix C: Code of Professional Responsibility of the Official Court Interpreters of the United States Courts. In Fundamentals of court interpretation, (pp. 585-586). Durham, N.C.: Carolina Academic Press.

Gerver, David. 1974. “The effects of noise on the performance of simultaneous interpreters: Accuracy of performance.” Acta Psychologica. 38:159-167.

Gile, Daniel. 1995. “Fidelity assessment in consecutive interpretation: An experiment”. Target. 7.1:151-164.

McIlvaine Parsons, H. 1978. “Human factors approach to simultaneous interpretation.” In David Gerver and H. Wallace Sinaiko, eds. Language interpretation and communication (pp. 315-321). New York: Plenum. Michelsen, Patricia. 1992. “Court interpreting.” The Court Management & Administration Report. 3.10-16.

Moser-Mercer, Barbara, Alexander Künzli and Marina Korac. 1996. “Prolonged turns in interpreting: Effects on quality, physiological and psychological stress”. Working Paper. University of Geneva, École de Traduction et d’Interprétation. Unpublished manuscript.