A New Approach to Technical Translation Teaching
3. The new approach
3.1 Source text selection on the basis of empirical market studies
The source texts used in technical translation exercises are selected so that they reflect the most typical types of texts (in terms of Textsorte, Texttyp and subject), i.e.
the teacher uses texts which are typical for translation practice in that particular lan
guage pair and direction of translation. Obviously, the translation of an automotive user manual is more typical in the direction German-Portuguese than in Portuguese- German. To know what is “typical”, beyond the horizon of an individual translator, it was necessary to conduct some market research. With regard to the German transla
tion market, many institutes are now using the results of my market surveys of 1989 and 1992 (Schmitt 1990a, 1990b, 1990c, 1993).
These surveys confirmed our expectations that what is relevant in translation prac
tice is also governed by the professional status of the individual translator. A freelance translator’s scope of text categories and subjects varies from those of a translator work
ing in an in-house environment, such as the translation service of a corporation or government agency. Considering the trend towards outsourcing of translation ser
vices, it seems wise to shift the focus of translation exercises to the types of texts and subjects that govern the work of freelancers and translation agencies. In the case of the language pair German and English, for example, this means clearly that manuals in the fields of mechanical engineering (including automotive) and electronics/com
puters should be given the most attention.
In the course of studies, the technical character of the text samples will gradually increase. In their first technical translation exercises students would learn how to handle consumer-oriented product documentation such as car owner’s manuals or simple software manuals. Expert-oriented texts, such as service manuals, require
Peter A. Schmitt
much more technical and terminological knowledge, have different textual chacteris- tics and are covered in later courses. A typical example for texts translated at the end of translation studies are patents, due to their combined nature which adds a legal quality to a technical (e.g. engineering-related) subject. The reasons for selecting a specific text as a source text in a technical translation exercise as well as the features of this text and its corresponding texts in the target culture are explained in a brief
ing during the first session.
3.2 Computer-based training (CBT)
Since text-related technical knowledge is indispensable for any technical transla
tor, it is a major task to convey that knowledge to our students, at least to the extent necessary for technical translation exercises. In addition to the traditional teaching approaches (classroom teaching by a teacher, individual studying using textbooks), which of course are still useful, the current wave of electronic reference on CD-ROM also brought us new teaching possibilities. The scope ranges from published CD- ROMs to specific, custom-made computer-based training systems or CRTs. The for
mer are the modern equivalent of textbooks and explain, more or less interactively and attractively, for instance, how things work (be it the human body, computer hard
ware, software or a car). Other examples are language trainers. The latter are programs specifically produced for translation studies. A typical example and may be the first of its kind is the program developed by Doris Ensinger and Willy Neunzig of the University of Barcelona. It is intended for Spanish German-learners and translation students and makes its user familiar with the political system in Germany.
Another example is the computer-aided translation teaching system, called CATTSY by its developer Lothar Schmiedel of the University of Leipzig. Basically, it consists of a set of typical source text examples which cover a wide spectrum of text types and subjects. A sophisticated system of hyperlinks connects text elements to a wealth of background information, including parallel texts, explanations of linguis
tic and theoretical nature, factual information, and access to technical background and terminology. Even scholarly articles on text-related phenomena can be accessed by a mere mouse click. Both systems will be demonstrated at the conference in Leip
zig next week. In the light of budget problems and current efforts towards a leaner and more efficient university it seems worthwhile to consider ways to deploy teach
ing manpower where it is indispensable and to use self-learning approaches wherever possible and didactically justifiable.
3.3 State-of-the-art translation tools
In 1991, Christopher Marsh reported about his experience with the use of per
sonal computers in translation exercises at the University of Innsbruck (Marsh 1991).
One year later, a similar teaching concept was implemented at the University of Mainz/Germersheim. According to this concept, the source text is translated by the students as a “homework” assignment, as usual, deploying all tools a professional translator would use, such as a computer work place with access to translation-related software. At the BASIC in Germersheim, about 50 networked state-of-the-art PCs with access to the FASIC’s inhouse terminology databank FASTERM and the world
Section 2. LSP and Translation
wide web are available to students. Taking into account that an increasing number of students have their own personal computers in their dens, 50 workplaces is no bot
tleneck despite the institute’s 2,300 students. The students bring their Target text to the next session one week later, either as a printout or, preferably, on diskette. The printout approach turned out to be more efficient when groups were larger than ten.
Technical translation exercises take place in a specially equipped class room. In this scenario, only one PC workplace is necessary, located next to the teacher’s posi
tion. The computer monitor is hooked to a data display or beamer to project the screen as large and brightly as possible. Students’ desks are arranged so that they face the projection screen of the workplace PC. The teacher is positioned so that she or he can face the students as well as the computer screen and its wall projection, without obstructing the light path of the projector (therefore, a ceiling-mounted beamer is preferable). Clarity and brightness of the projected screen image are a cost factor;
in Germersheim, a VGA-color data display is used in conjunction with a 400-watts xenon projector whose bulb costs about 500 DM and has a life of about 500 hrs, turning dimmer and less usable at the end. In any case the room must be dimmable to improve contrast and legibility of the projected image, with sufficient light available so that students can still write and read their notes. In the prototypical Germersheim layout, the PC is networked and provides access to all hardware and software resources that might be useful for a translator (for details refer to Schmitt 1994a, 1994b), in
cluding a CD-ROM tower loaded with electronic reference.
3.4 Role games
In each session, three students are simultaneously and actively involved. One stu
dent acts as a system operator, usually for an entire session, sits at the PC workplace and is in charge for physically producing the target text. In an ideal group of twelve and over a typical twelve-week semester, each student can assume this role at least once. The ability to handle a PC and a word processor is a prerequisite for partici
pating in such exercises, it can be acquired in separate courses.
A second student assumes the translator’s role, the assumption being that the translator has produced a translation for a client (which may be the translation’s end user or a translation agency).The translator student either gives the system-operator student a diskette with the target text file she or he has prepared for this session or dictates one segment (usually a paragraph) of the target text to the system operator who keys it in. If on diskette, the text is loaded into the word processor’s target doc
ument. To avoid delays due to conversion (and file format conversion problems) dur
ing the session, students’ file contributions should have the same format (such as RTF or plain ASCII or ANSI). By the way, dictating a text to a typist is in itself some
thing worth learning, as initially students tend to speak not clearly or loudly enough, omit punctuation and layout information.
A third student plays the role of proofreader. Creating awareness of the impor
tance of translation quality control in the light of current boundary conditions of the language industry, such as product liability (Schmitt 1996: 50) and certification according to ISO 9000, is one of the key goals of translation exercises.
Peter A. Schmitt
3.5 Quality control
With reference to the quality assurance procedures used in industry, such as those of the Translation Center of IBM in Germany (Schmitt 1997), the teacher encourages the proofreader to proceed in three steps:
First, the translator’s target text is checked for obvious formal deficiencies, such as missing text segments, violations of grammar, spelling and punctuation. The proof
reader points out whatever he or she considers as a defect. The translator has now an opportunity to defend his or her version; the proofreader in turn should be able to explain his or her position. Besides the obvious effect of improving the target text, this discussion enhances the students’ ability to talk professionally about their job, using their meta-language tools, and to argue carefully, getting their point across without destroying human relations in the office. Once this is settled the proofreader asks the system operator to edit the text according to instructions. When the proof
reading student says that the target text is now formally acceptable, the teacher asks the other students in the group whether they approve of this or whether there are still mistakes that need to be eliminated. As everybody can easily watch the progress of text improvements on the screen, instead of having to rely on memory and notes, students tend to be much more active and committed in this step of the process than in traditional translation exercises, coming forward with suggestions that lead to fur
ther improvements. The teacher’s role is that of a moderator who controls the course of events and draws the students’ attention to problem areas, explaining the positive and negative effects of certain text changes and variants. Only if a problem remains unnoticed or unresolved by the students the teacher will be called upon to make final corrections.
In the second step, the target text is checked for information correctness, the cri
terion being that the text should meet the minimum level of quality necessary to for
ward it to the client. This simulates a typical situation where meeting a deadline is more important than elegance of style. However, this criterion does account for ter
minological accuracy, e.g. use of the client’s corporate language and not the termi
nology of competitors. It also encompasses observation of certain elementary text features, such as use of non-personal style in German service manuals. The proce
dure is the same as in the first step.
The third step postulates that there is sufficient slack in the project schedule to do whatever is considered beneficial to the document. This covers any aspect of good technical writing, such as clear theme-rheme progression, use of examples which are customized to the target readership, as well as minor or even major changes to the text’s micro or macro structure. Here, the target text may be changed radically or modified in small iterations - good writing means re-writing. Depending on the nature of the text and translation exercise, this step may also include formatting aspects such as layout and typesetting. Otherwise, the overall procedure is again the same as in the preceding steps.
Wherever possible, problems arising during the process are resolved by accessing available online resources, such as databases, to acquaint the students with the limits and possibilities of these media (for details refer to Fischer/Freigang/Mayer/Reinke 1994). Once consensus has been reached, the current text segment is considered fin
Section 2. LSP and Translation
ished, and a new team of translator, proofreader and system operator tackles the next text segment.
The resulting target text is either printed out at the end of each session, or, as this might create an excessive delay with large groups, it is handed out the following week.
A different approach and scenario is used in some other places, such as in Leipzig and Erlangen. There, every student has a PC workplace of his or her own, and by means of a special teaching network and switchboard, the teacher monitors the trans
lation procedure of each student in turn.
3.6 Realistic translation scenarios in examinations
For various reasons, the translation procedure employed in final examinations remained largely unaffected by any technological change. The prevailing picture is still a lecture hall filled with candidates crouching over printed source texts and busily penwriting their translations on sheets of paper. A new approach is now being implemented in Leipzig. As of 1997, technical translation examinations will be held in one of the computer rooms. Each candidate has a translator’s workbench at its disposal, networked, with a defined spectrum of resources, including electronic dic
tionaries, creating absolutely equal opportunities for everybody. The candidates may use what-ever they feel necessary to produce a target text printout of the defined quality within the defined time frame.
This approach seems rather bold as compared to examination conditions at other institutions, some of which do not even allow the use of dictionaries, not to mention computers. The new approach postulates that translations produced in examinations at the end of translation studies are not to be treated as vocabulary tests. Instead, candidates are to show their ability to efficiently use professional tools to achieve a defined goal, both in terms of quality and time.
Summary
In conclusion, let me summarize the following aspects which I consider to be the most essential:
• Source texts selected on the basis of empirical market studies achieve a maxi
mum didactic benefit with a minimum of effort.
• Proper usage of CBTs for individual learning relieves the teaching staff from repetitive routine and makes personnel available for more demanding teaching tasks.
• Computer-based translation tools in translation courses familiarize students with state-of-the-art working processes.
• Wall projection of the target text increases the students’ awareness of formal aspects such as proper spelling, punctuation, typography and layout, as well as of transphrastic features such as text coherence, sequencing of information, theme-rheme progression, and consistency of terminology.
• Role games involving a translator, a proofreader and a computer system opera
tor create a brain-storming atmosphere which stimulates in-class student involve
ment and improves the students’ teamworking ability.
Peter A. Schmitt
• In-process evaluation, proofreading and editing enhance the students’ commit
ment to translation quality.
• A defined set of evaluation criteria and a transparent ranking system improve the acceptability of grades and the therapeutical effect of translation tests.
• Realistic translation scenarios in examinations, providing defined access to up- to-date hardware, software and electronic reference, create equal opportunities for every student and allow to assess the students’ ability to work under profes
sional boundary conditions.
Looking back on about five years of experience with our new approach to tech
nical translation exercises we believe that it tremendously improves the student’s awareness of the often delicate features which may make all the difference between a good and a better text. When competing with automatic translation systems, the ability to produce high-quality translations is, so to speak, the USP (unique selling proposition) of human translators.
References
Fischer, I. & Freigang, K. & Mayer, F. & Reinke, U. (eds.) 1994. Sprachdatenverarbeitung für Übersetzer und Dolmetscher. Aktén des Symposiums zum Abschlufi des Saarbrücker Modell- versuchs, 28.129. September 1992. Reihe Studien zu Sprache und Technik, Bd. 5. Hil- desheim: Olms.
Honig, H. 1995. Konstruktives Übersetzen. Tübingen: Stauffenburg.
Kussmaul, R 1995. Training the Translator. Amsterdam: Benjamins.
Marsh, Ch. 1991. Textverarbeitung im Übersetzungsunterricht. Lebende Sprachen 3: 97-98.
Schmitt, RA. 1990a. Die Berufspraxis dér Übersetzer. Eine Umfrageanalyse. Berichtssonderheft des Bundesverbandes der Dolmetscher und Übersetzer. Bonn: BDÜ.
Schmitt, P.A. 1990b. Was übersetzen Übersetzer? - Eine Umfrage. Lebende Sprachen 3: 97- 106.
Schmitt, P.A. 1990c. Übersetzer/Übersetzerin, Dolmetscher/Dolmetscherin. Eine Umfrage
analyse über die Berufspraxis. Informationen für die Beratungs- und Vermittlungsdienste der Bundesanstalt für Arbeit 12: 517-533.
Schmitt, P.A. 1993. Der Translationsbedarf in Deutschland. Ergebnisse einer Umfrage. Mit-
teilungs blatt für Dolmetscher und Übersetzer MDÜ 4: 3-10.
Schmitt, P.A. 1994a. Computereinsatz im Übersetzungswesen. Ein Überblick. In: Beck, U. &
Sommer, W. (Hrsg.) LEARNTEC 1993: Europáischer Kongrefi für Bildungstechnologie und betriebliche Bildung. Tagungsband. Berlin: Springer, 567-590.
Schmitt, P.A. 1994b. Translationsrelevante Software. BDÜ-Mitteilungsblatt für Dolmetscher und Übersetzer (MDÜ) 3, 3-11. An updated version may be downloaded via the Internet under the WorldWideWeb URL http://www.fask.uni-mainz.de/ze/term/cat-pas.html
Schmitt, P.A. 1996. Warnhinweise in deutschen und englischen Anleitungen. Lebende Sprachen 2: 49-57.
Schmitt, P.A. 1997. Evaluierung von Fachübersetzungen. Schmidt, H. & Wot jak, G. (eds.)
Modelle der Translation. Festschrift zum 65. Geburtstag von Albrecht Neubert. Tübingen: Narr.