Article for translators: Time Pressure and Dictionary Availability

Home

Time Pressure and Dictionary Availability

By Leonardo Jordão Coelho

leojbrazil@hotmail.com

UNIVERSIDADE FEDERAL DE MINAS GERAIS
FACULDADE DE LETRAS
DEPARTAMENTO DE LETRAS ANGLO-GERMÂNICAS

Become a Member of TranslationDirectory.com at Just 4 EUR/Month (Paid Yearly)

Advertisements:

Monografia final
apresentada ao curso de Especialização
em Inglês da Universidade Federal de
Minas Gerais, como parte do requisito
para obtenção do Certificado de
Especialista em Inglês.

Orientador: Prof. Dr. Fábio Alves

To my mom Maria do Carmo Jordão,
who has always encouraged my interest
in academic studies, saying it is possible.
To my dad Miramar Coelho and to God
for their company.

Summary

1 - Introduction

2 - "Time Pressure in Translation"
Astrid Jensen

3 - "A study of use of dictionaries in Danish-English translation"
Inge Livbjerg and Inger Mess

4 - Conclusion

Introduction

Electronic documentation has become an essential tool in the area of translation studies research. Needless to say, psycholinguistic methods of verbal reporting (Think-Aloud Protocols - TAPs) have also been efficiently used to conduct scientific experiments in this field. This paper is concerned with a summary of two articles, which were published by the Copenhagen Business School in its publication "Probing the process in translation: methods and results" in 1999. The two articles were chosen because both of them are concerned with two essential issues that worry most translators: time pressure and dictionary availability.

The first article "Time Pressure in Translation' by Astrid Jensen focuses upon the results of an experiment carried out to confirm if and how time constraints interfere in three groups of translators at different within three different periods. The paper also outlines the processes and strategies applied by these three groups to cope with time pressure. In this respect, the study investigates if both professional translators and non-translators applied coping tactics when solving problems related to dead-lines and gaps in their own linguistic knowledge. Electronic documentation from the Translog program and its think-aloud protocols were also used as a basis for this paper.

The second article "A study of the use of dictionaries in Danish-English translation" by Inge Livbjerg and Inger M. Mees investigates how the use of dictionaries might influence the process of translation of a text. For the experimental study, the researchers compared ten different translations done by five competent post-graduate translation students. These subjects were asked to translate a general-purpose text from their Danish mother tongue into the English language using the Translog program. The think-aloud protocols were also analysed by the researchers. The first time, no dictionaries were available. After five translations were done, the researchers asked the translators to review their work, now with the possibility of looking words up in dictionaries if they wished to do so.

The focus of the present work is to summarise the main ideas, points and concepts that were provided in both articles.

Time Pressure in Translation

Astrid Jensen

The article looks at findings from the process-oriented part of a PhD project to develop and test hypotheses about how translators cope with time pressure in newspaper translations, following the modern trend of studying the mental processes in translation based on verbal self-reports of cognitive processes.

The methodology employs experiments using the think-aloud protocol (TAP) method of data elicitation. In this regard, translators translate a text and verbalise as much of their thoughts as possible. It is assumed that information is stored in memory with different capacities: a short-term memory with limited capacity and for limited duration, and a long-term memory with large capacity. Information recently obtained is kept in short memory, where it can be processed and reported on. However, only conscious processes can be reported: any automised skills may by-pass short-term memory. Thus, it is not always possible to infer all of professional translators' thought processes from their verbalisation. Another problem with concurrent reports is the extra cognitive load this represents; the person may need to stop verbalising to carry out the main task.

This introspective method was supplemented with computer logging. All translations were done using Translog and concurrent verbal reporting. (Translog records all keyboard activity and saves it in a log-file.)

Coping strategies are mentioned as being used by experienced translators as a mean to reach the goals set by the translation task.

The purpose of the study was to attempt to locate affects of time pressure, to identify processes and strategies used by translators and non-translators, and to verify whether coping tactics would be used by both groups under time pressure and with insufficient background knowledge. In particular, the questions to be answered were:

1. "Does time pressure impose a restriction on problem-solving activities?"

2. "Do different groups apply different strategies to cope with time pressure, and what are the indications of coping tactics?"

The study was based on experiments carried out in the 1997. The informants represented different levels of proficiency in translation; four were professional translators and two were "educated laymen", who use English as a working language. All translated three texts, each on a different topic and from a different source, from English into Dutch, their mother tongue. Time limits of 15, 20 and 30 minutes were set, and the texts were to be translated as if to appear in a quality Dutch newspaper. The average length of the texts was 120 words. The participants could use dictionaries at will and worked in their own offices.

The data analysed was composed of the text, divided into segments, each of one complete sentence, the participants' verbalisations, the Translog data and statistical data. This latter comprised the number of words in the segment, the number of letters, the time taken for translation, for revision and total time taken.

A 4-second pause was chosen as an indicator of a potential problem-solving activity. This suppressed delays due to differences in typing speed. Some pauses were identified as dictionary look-ups, other as comments on problems, memory search or word associations. However, since not all pauses were commented on, they were grouped all together as pauses suggesting problem-solving. A "coding potential" was made to identify processes that took place during translation, which could indicate problem-solving activities. The main source used was Translog (except for dictionary look-ups, which were identified from the TAP).

The specification of translation strategies included borrowing, literal translation, paraphrasing, adaptation, and reduction. Coping tactics were identified as instant naturalisation, transcoding, reconstruction using context, generalisation and omission.

Patterns of common behaviour within groups were identified once it was decided to divide the translators into two separate groups: those with 1-3 years experience, or young professionals, and those with 8-10 years experience, or expert group. Therefore, for the purpose of presenting results, the participants were divided into three groups: non-translators, young professionals and the experts.

The participants all read the text before translating it. They went through the text in a basically linear fashion, translating text elements and, in some cases, editing them. When they had finished, they reread their translations in most cases.

Analysing the results, there was a 14% increase in the number of problem-solving activities and a 21% increase in the number of keystroke when the limit was raised from 15 to 30 minutes. However, there was a 40 % decrease in the speed of typing under the increased time limit, although this may be partly due to increase dictionary look-ups and problem-solving pauses.

As regards the use of dictionaries, there was a clear difference among the three groups. The non-translators used dictionaries most, followed by the young professionals. The experts had approximately half as many look-ups as the young professionals.

More editing took place in the group of young professionals than in the other groups. Non-translators had only about half as many corrections after completing the translation as the experts; who had the same number of corrections during revision as during translation. It appears that the time constraint was felt most by non-translators.

The number of problem-solving activities was highest for the young professionals and lowest for the experts (a difference of 30 %). This suggests an increase in automation of thought processes for the latter group.

The article also gives examples from the translations of each of the translation strategies mentioned earlier. A statistical analysis of the coping tactics used is presented in tabular form. It shows that the main coping tactics used by non-translators were transcoding and omission. On the other hand, the main coping tactics used by professionals were reconstruction and generalisation.

In short, it was found that there were substantial differences not only in the way non-translators and translators dealt with time pressure, but also considerable differences between the young professionals and the experts. All groups, decreasing dramatically with experience used coping tactics. However, it was considered that it would be premature to conclude that the use of coping tactics was caused by time pressure alone. A continuation of the study is planned to ascertain if time pressure plus lack of relevant background knowledge is what motivates the use of coping tactics.

A study of use of dictionaries in Danish-English translation

Inge Livbjerg and Inger M. Mees

The article reports the results of a study carried out to compare how students' translation process without dictionaries differed from their approach when dictionaries could be consulted. Five post-graduate translation students was asked to translate a newspaper article of 126 words from Danish to English while thinking aloud, and spending as much time on the task as they felt necessary. Translog was also used to register the process.

The text chosen to be translated was a topical one with a certain degree of complexity, in which there were several types of problem areas, for instance, metaphors, collocations and potentially ambiguous expressions. The intention was to see to what extent the subjects could arrive at good solutions without dictionaries, and also to see if subsequent access to dictionaries would result in changes in the translations.

The text was first translated without the possibility of consulting dictionaries. The completed translations were saved. After a break, the subjects were asked to look at their translations again to see if they would like to change anything. This time they had access to reference books: a usage book, a Danish-English dictionary and a monolingual English dictionary. However, they were not specifically asked to make use of these. The final result was also saved, so that for each translator there were two products, one without the possibility of consulting dictionaries and one with. The researches were in a different room from the subjects and could see them through a glass panel and hear them via an audio link. Whenever a translator consulted a reference work, the book and page numbers were noted by the researches.

A professional, bilingual translator provided a model translation of the passage, a native speaker assessed the translations as texts, and two experienced Danish translators assessed them as translations. These evaluations determined whether the product had improved, deteriorated or remained stable.

The study compared the first spontaneous solution produced by the subjects and the final suggestion opted for at the end of the stage 1, as well as between the end product of stage 1 and the end product of stage 2. It should be noted that it was not always easy to determine exactly what the first spontaneous solution was. In one hand, an error could be kept, corrected, or changed into a different error; still, a correct solution could be kept or changed into an error or a different correct solution. The problems analysed were those identified as such by the subjects. Moreover, it was important to use Translog and think aloud protocols so that it was possible to have access to the whole translation process. The researchers needed to find out in what point a unit was identified as problematical and if it had been the dictionary or some other process or strategy that had been used to solve the problem.

Altogether, subjects commented on 76 problems. However, a large number of these were common to different subjects: there were only 23 different units in all, which shows a high degree of consensus as to what constitutes a problem.

In stage 1, from the first solution to the one adopted, 48 did not change and 16 were changed either from a correct solution to another correct solution or from an error to a different error. Only 2 correct solutions were changed to errors and 8 errors were corrected.

In stage 2, from the solution adopted at the end of stage 1 to the one adopted after the possibility of consulting dictionaries, 55 did not change; 9 were changed from one correct solution to another or from one error to another. Further more, 3 correct solutions were changed to errors and 7 errors were corrected.

Analysing the whole process, in 56 instances (75.7%) there were no major changes in the quality of the product, that is, a solution which was correct continued correct or changed to a different correct solution, and one which was an error remained an error or changed to another error.

In conclusion, if subjects should retain their first spontaneous choice it would give appropriate solutions in only 52.7% of cases. The process without dictionaries raises this percentage to 60.8% and that with dictionaries to 66.2%. So both have a positive contribution.

Dictionaries were used by subjects to look up between 5 and 14 units, the actual number of consulting being from 10 to 27 since some units involved more than one consultation. Use was made of dictionaries to solve gaps in subjects' vocabulary, to check collocations, for fear of false friends and sometimes because of problems of reception. The original text contained at least 5 items for which there were reception problems, despite being the subjects' native language.

The study also showed that subjects spent a long time on problems for which they had a solution from the start, although this could be a result of the conditions of the experiment. It is not considered, however, that restricting time or avoiding the use of dictionaries is cost-effective, since the subject whose final products were regarded as best by all evaluators was the one who took most time and had most dictionary look ups. Analysing the above, it was concluded that students of translation should be taught to make better use of dictionaries, as well as strategies such as paraphrasing or omission. It is also suggested that organising translation workshops is the best way to further improve the performance of translation students.

Conclusion

After analysing both articles, it can be concluded that expert translators and non-translators actually differ in dealing with the same translation task. Indeed, such remarkable difference lies mainly in what specific strategies are used.

On one hand, experts seemed to have fewer problem-solving activities and did not feel time constraint as much as non-translators did. Actually, as regarding time constraints, there were experts who felt the time available was much more than enough to accomplish the task. In general, they were used to accomplishing such a task in much less time in real work life. On the other hand, non-translators seemed to favour a word-by-word or literal translation and did spend a great deal of time pondering over, or checking words to which they already had a solution from the very beginning. Summing up, non-translators were the ones who most felt time constraint.

In other words, non-translators seemed to be unaware of their own linguistic competence and/or felt insecure about their spontaneous translation. To a certain extent, they depended much more on the linguistic structure of the text while translating and were the ones who most had dictionary look ups when compared with the experts.

Regarding dictionary use, there were cases in which they were used not only to correct errors, but also to change correct solutions into errors. In sort, the second article shows that some subjects did not distinguish "between solutions that can be found in dictionaries and solutions that involve the use of other communication strategies such as paraphrasing or omission".

Based on these assumptions, the conclusion drawn by the present study is that translation students must be taught to make a better use of dictionaries and to master abilities to exploit co-textual and contextual clues. Considering that paraphrasing, generalisation and omission were the strategies most used by the expert group and that such strategies seems to increase with experience; mastering them must always be the goal for translation students in order to help them to reconstruct another text in another language - that is what translation is all about.

Bibliography

CIRANKA, Lúcia Furtado de Mendonça e SOUZA, Vânia Pinheiro de. Orientação para normalização de trabalhos acadêmicos. Ed. EDUFJF. Juiz de Fora, 1993.

COLLINS COBUILD ENGLISH GRAMMAR. London & Glasgow: Collins, 1990.

COLLINS COBUILD ENGLISH LANGUAGE DICTIONARY. London & Glasgow: Collins, 1990.

COPENHAGEN STUDIES IN LANGUAGE 24. Probing the process in translation: methods and results. Copenhage Business School. Samfundslitteratur, 1999.

DICIONÁRIO OXFORD ESCOLAR. Para estudantes brasileiros de Inglês. Oxford University Press, 2001.

LONGMAN DICTIONARY OF CONTEMPORARY ENGLISH. Harlow: Longman, 1986.

NOVO DICIONÁRIO AURÉLIO DA LÍNGUA PORTUGUESA. 2. ed. ver. amp. Rio de Janeiro: Nova Fronteira, 1986.

OXFORD ADVANCED LEARNER'S DICTIONARY. Sixth Edition. Oxford University Press, 2001.

SEVERINO, Joaquim Antônio. Metodologia do trabalho científico. Diretrizes para o trabalho didático-científico na Universidade. Ed. Cortês. São Paulo, 1982.

SWAN, Michael. Practical English Usage. Fourteenth Edition. Oxford University Press, 2002.

THE BBI DICTIONARY OF ENGLISH WORD COMBINATIONS. Amsterdam and Philadelphia: John Benjamins Publishing Co., 1997.