CC BY
9Journal of Siberian Federal University. Humanities & Social Sciences 12 (2016 9) 2913-2916
УДК 81'25
The Role of Visual Context for Stylistic Differences in Students' Translations of User Interface
Stanislav G. Bogdanov*
New Bulgarian University 21 Montevideo Str., Sofia 1618, Bulgaria
Received 26.11.15, received in revised form 17.03.16, accepted 27.09.16
This paper reports on a case study aimed to investigate the role of visual context for stylistic differences in students' UI translations. Data from the two groups revealed stylistic differences and a tendency for a more consistent use of already conventionalized vocabulary on UI elements by the experimental group.
Keywords: audio-visual translation, user interface translation, visual context, localization. DOI: 10.17516/1997-1370-2016-9-12-2913-2916. Research area: philology.
Introduction
Ivan Turgenev wrote in Fathers and Sons in 1862, "The drawing shows me at one glance what might be spread over ten pages in a book" (Рисунок наглядно представит мне то, что в книге изложено на целых десяти страницах). This old adage reworded as "A picture is worth a thousand words", is all the more valid today in software and web-based user interface (UI) translation.
Literature review
Dual coding theory (Clark, J. M. & Paivio, A. 1991) posits that an appropriate visual clue which accompanies textual information greatly aids comprehension. It is common practice for software companies and translation agencies to send out to translators and localization experts alphabetical
lists of words and phrases devoid of visual context with the wrong assumption that one or two words are ever so easy to translate. The issue aggravates with target languages like Bulgarian where there are no established conventions yet and where two forms of addressing the user exist.
While research findings in support of the Dual coding theory are abundant, very little research has been carried out into the impact of visual context on UI translation. Therefore, the aim of this small-scale study is to investigate the differences that visual context produces in translating software and web-based UI.
The study
1. Research questions
The study analyzed the role of visual context for differences in style and register of students'
© Siberian Federal University. All rights reserved
* Corresponding author E-mail address: stanbogdanov@nbu.bg
- 2913 -
UI translation. It aimed to answer the following questions:
1. To what extent does visual context influence the style and register in students' translations of UI?
2. What differences are observed in students' UI translations with and without visual context provided for the UI translation?
2. Participants
The study was conducted at the Department of English Studies of New Bulgarian University. The programme equips students with practical skills through hands-on experience in translation of various types.
Ten graduate students from one Translation in Localization class participated in this study. The participants' English language competence is C1-C2 (CEFR). They met for 90 min once in a week. Their average age was 35.8 (age range 24-56 years). Everyone had used a computer before with an average of 15.8 years of usage experience.
3. Design and procedure
The present study is based on analyzing the linguistic output produced by students in the class tasks. They included web-based and software UI translation.
The source text contains 3883 words, 991 of which are unique words. The target text contained on average 3759 words, approximately 1126 of which are unique words.
Two groups of students translated a CMS plug-in. The experimental group had screenshots while the control group had the source text only, without any visual support.
Results and discussion
Analysis of the students' translations revealed the following differences between the experimental and the control group.
1. The amount of variance in translations, i.e. the output of the control group was bigger, while the experimental group was more consistent in using already conventionalized vocabulary.
2. Nominalization: a lot more instances of nominalization were observed in the control group than in the experimental group. A possible explanation is that the students tried to compensate for the lack of clarity of communication due to the lack of clear context.
3. Redundancy and explication were the other features that were observed in the control group, e.g. (Playing ... = В момента звучи) -explication compensating for Progressive aspect
4. More complex language use in the control group.
5. By definition UI translation does not use complex style but rather aims at simple style. Simplification is also observed although with attempts to retain complexity so as to be true to the original.
The author was not interested in the students' mistakes; more important was why it happened. What was observed was only at the surface level: to help understand the reasons and processes, to go back to communication and language deeper analysis is needed.
Communication involves ostension (linguistic or not) and inference. The inferential process involved in communication is the creation of a context in which an ostensive act finds its relevance.
The role of all aspects of language use in communication is to constrain the inferential process, to help the addressee construct a context in which the communicator's ostensive act can be seen to be relevant. So, it is not the context that disambiguates language, but language that disambiguates the context of interpretation. And the more explicit the
linguistic form, the more constrained the addressee is in constructing the context of interpretation. (LaPolla, 2003)
Consequently, first, communication can take place without any language involved. That's not new. Second, language does restrict the context of interpretation of the message.
The inferences involved in interpretations are guesses at what the intended message might be. Language reduces the number of assumptions that could potentially be part of the interpretation context; and the more explicit the utterance, the more constrained the interpretation.
How does that relate to UI translation? Interface, Internet, software and all technology and computer mediated communication is mostly performed through a visual medium. Therefore, users are communicating visually. Language is present to restrict the user of the UI in interpreting the context as intended by the communicator. For example, an empty button may signify both "upload" and "download".
So, both visual context and language together are used to communicate to the user. Therefore, the translator needs visual context so as to be able to 1) interpret the message correctly, and 2) to translate it correctly for the target user in the other language and culture. And what is more, this message must mean the same to anybody using the interface.
Conclusions
Going back to the results from the current experiment - the amount of variance - the students in the control group had no visual clues and there were many possible interpretations as they did not have one of the two components of communication. The students made sure they passed on the message to the user, which lead to redundancy and explication so as to compensate for the lack of visual context in the situation where co-text was also confusing.
Regarding nominalization and more complex language use, having no visual clues lead to a higher level of formality than needed. The following example illustrates some of the difficulties:
e.g. "download —> свали / свалете —> изтегли / изтеглете"
The above is observed in languages without an infinitive form. Russian has an infinitive form and easily resorts to "загрузка" or "скачать". In Bulgarian, this is also done through nominalization, which makes it impersonal, but also very clumsy.
How to teach the translator as a localization expert?
First, we need to educate the translator to demand visuals so that we start educating the agencies and the customers. If visuals are not available, we must teach students to visualize.
Here the author proposes a didactic model -a 3-step approach, which the author also uses in a subtitle translation graduate course.
The model
1. Analyze the text, help the students visualize, discuss where and how this could be located in the design: if it is a button, a menu, help text; what part it is for the administrator, what part it is for the front user, who the target audience are; culture specific aspects. Then encourage students to think as developers and designers.
2. Students translate
3. Discussion of students' translations -students justify their translation solutions and think of improvements. What follows is a discussion on clarity, compression, consistency and avoidance of synonymy in UI.
Drawing from the results of the current small-scale investigation, attention should be
paid to the following when training UI translators
The above model is to be tested empirically in further studies. Lack of context can affect the meaning. Therefore, students should be taught to navigate the "co-text world" presented in the alphabetical order in UI translation; to make mental links to the context of each phrase/sentence and should be encouraged to visualize the context.
as localization experts:
• Formality vs. Informality
• Simple vs. Complex Style
• Fronting
• Repetition and Variation
• Redundancy
• Passive vs. Active style
• Nominalization and Verbalization
References
Clark, J. M. & Paivio, A. (1991) Dual coding theory and education. Educational Psychology Review, 3(3), 149-210.
LaPolla, R.J. (2003). Why languages differ: variation in the conventionalization of constraints of inference. In D. Bradley, R.J. LaPolla, B. Michailovsky, & G. Thurgood (Eds.), Language Variation: Papers on Variation and Change in the Sinosphere and in the Indosphere in Honour of James A. Matisoff (pp. 113-144). Canberra: Pacific Linguistics, Australian National University.
Стилистические различия
в студенческих переводах интерфейса пользователя: роль визуального контекста
Анализируются результаты исследования роли визуального контекста как фактора появления стилистических различий в студенческих переводах интерфейса пользователя. Данные переводов двух групп студентов продемонстрировали ряд стилистических различий, а также тенденцию к более последовательному использованию уже устоявшейся терминологии интерфейса пользователя.
Ключевые слова: аудиовизуальный перевод, перевод интерфейса пользователя, визуальный контекст, локализация.
Научная специальность: 10.00.00 - филологические науки.
С.Г. Богданов
Новый болгарский университет Болгария, София 1618, НБУ, ул. Монтевидео, 21
