1Translating is not a simple task. It has long been established that it consists of much more than the mere replacement of word (a) of language (A) by word (b) of language (B) (Horn-Helf 1999); or, as well-known German scholar Hans G. Hönig (1995: 13) put it:
- 1 „[…dass] der Übersetzungsvorgang mit dem Austauschen von Wörtern sowenig gemeinsam hat wie das Scha (...)
The translation process has as little in common with replacing words as playing chess does with pushing chess pieces around on the board. (Our translation)1
2This so-called problem of equivalence is discussed with first-year students in their “Introduction to translation” class and even individuals beyond the inner circle of the scientific and professional community are no longer fully ignorant of its implications.
3The same is not true for a different kind of translation complexity. As early as 1984, Holz-Mänttäri described the dependence of any translation task on the context for which the target text is produced. Nowadays, a translator’s job implies teamwork or co-operative work, the acquisition of projects and the organising of tasks. The translation process cannot be separated from its context, it is constantly influenced by (and draws on) the physical and technological means and “human resources” that surround it. Lately, such thinking has also found its way into scholarship. As an example, Gouadec (2007: 56f) has coined the term “translation service provision process” for his findings. The teaching of translation, however, is lagging behind. This may partly be due to practical hurdles, but it is also almost certainly rooted in the complexity of translating, as referred to by the term “expert character” in other works on the field (Risku 1998).
4In this article, a very brief overview will be given of research on translation as an expert activity narrowly tied to its concrete environment. I then outline the central role of computer aided translation tools (CAT tools) in this context. Next, I explain how the constructivist paradigm can be used as the basis for an adequate teaching method. Following this, I go on to illustrate how a constructivist approach can be put to use in the classroom through simulation. I then outline how I actually put the simulation into practice, and present data from two iterations of the respective Master-level course. The discussion focuses on the learning and motivational results of the approach, including the ability of students to cope with the complexity of the matter and method.
5Researchers dealing with the work of professional translators have come to call their subject “translation service provision process” (Gouadec 2007), “translation management” (Risku 2004) or “process of specialized communication” (Redaktionsprozess in German) (Schubert 2009). This underlines several characteristics of a professional translator’s activity: Firstly, present-day translators need to carry out many different tasks which are organisational or managerial in character. In smaller translation agencies these may occupy more of their time than translating (Kuznik & Verd 2010: 36). Secondly, what might appear as a single plain translation or editing job is frequently divided up between several individuals (Van Vaerenbergh & Schubert 2011: 20), who may be located on different continents. Thirdly, even when translators do translate, or work on other language-related tasks such as editing and correcting, this activity is strongly influenced by the work process it is embedded in (Risku 2004; 2009). This includes the material and technological aspects of the working environment as well as organisational factors and interaction with colleagues across departments. It adds to the complexity of the matter that the aforementioned elements are not only cumulative, but are also engaged in a “rather a constant and dynamic interaction” (Kuznik & Verd 2010: 26). Translating has therefore come to be conceptualised as a “situated” and collaborative activity (Risku 2004: 87f; Kuznik & Verd 2010; 26). Risku (1998: 89) has also argued that translating is an expert task, as marked by the need to solve complex problems. The type of problem-solving necessary for such tasks is, in turn, defined ex negativo as non-automated. This means that a high degree of flexibility in dealing with tasks and situations must be present so that a number of actions qualify as “expert”.
6All the abovementioned arguments stress that translating is strongly intertwined with the physical, technological and social context within which it takes place. Many of these aspects are mirrored in the functions of CAT tools, as the following section will explain.
7CAT tools are computer-based working environments tuned to the needs of the translation task. In former times, tools comprised two core functions: translation memory and terminology database. Translation memory means that a sentence, or clause, is stored together with the respective translation. Once the same sentence comes up for translation again, it is presented to the translator as a “ready-made” solution. Terminology databases can be considered as specialised glossaries, built up through the collection of terms that come up as the translator works on different projects. These terms can then be easily accessed within the CAT tool working environment. Both functions are very useful in controlling translation quality and making the translation process efficient. However, they are also known to cause quite a bit of collateral damage when incorrectly used. For example, finding a high percentage of matches in your translation memory (TM) will reduce translation time and cost. But a sentence wrongly matched or with a spelling error will pop up endlessly and is hardly ever corrected in an environment in which time is money and responsibility is infinitely diluted. The same is true for term bases. Figure 1 illustrates the two core functions.
Figure 1. Screenshot from the CAT tool Across – Translator’s desk
The translator’s workspace, including translation memory (displayed on the bottom of the screen) and access to the terminology database (on the right hand side of the screen)
- 2 For further information on the software see Across. 2010. Across at a Glance v5.0 SP 1. http://www. (...)
8Present-day CAT tools offer functions beyond the mere storage and recall of sentence pairs and terms, for example customer-specific management of such language resources, semi-automated workflow control, and basic communication channels. This is also true of Across, the specific software used for the simulation classroom presented below2. Figure 2 shows another screenshot from Across. The multitude of icons and buttons conveys an impression of the multitude of possible uses. The buttons on the left give access to the different types of tasks the software supports.
Figure 2. Screenshot from the CAT tool Across – Overview of functions
Functions covering the whole range from translation to project management tasks, as symbolized by the icons
9It is thus no exaggeration to say that present-day CAT tools are designed to help with the whole of the translation service provision process. In a business-driven environment, the tools serve to manage linguistic resources. They are also used for project management. And it is not rare that all communication between a translation project manager and project team of freelance translators and proof-readers has the tool as its only channel, as illustrated by Figures 3 & 4.
Figure 3. Screenshot from the CAT tool Across – Project manager’s view
Tasks related to a translation project (upper half of the screen) can be assigned to different team members (lower half of the screen)
Figure 4. Screen shot from the CAT tool Across – Team member’s view
A translator is informed that a task has been assigned to him or her through an automated system message
10Given the practical relevance of the functions discussed above, it is essential that these be taught – and successfully – as part of any translator training programme. However, taming and trimming a situated expert activity such as translation for didactic purposes is a considerable challenge. This is aggravated by a traditional divide between text-related and methodological courses at German academic institutions. However, the real causes lie deeper, and can only be overcome by taking into consideration the expert and situated characteristics of professional translating.
11If one looks at the curricula of translation training institutions in Germany, there are two types of courses; those strictly focused on teaching linguistic competence and translation (at my institution, Bachelor students have to complete 50 credit points and Master students 26 credit points in this field in order to graduate) and those requiring that students carry out authentic, practice-oriented translation projects. Bachelor students at our institution have to earn 5 ECTS from such a course, while the same type of courses is not mandatory at the Master level. What matters most, however, is not the limited amount of time spent on learning the unavoidable corollaries of translating in a professional environment, it is the manifest gap that students encounter between courses dealing roughly with “language problems” and those intended to convey the look and feel of the “real world”.
12Judging by their comments on, and behaviour in relation to translation projects, many students appear to be overwhelmed by the transition from text work to the complexities of real tasks. What is needed is an intermediate type of instructional setting. Students must be given tasks that are complex and challenging enough to help them grow into the reality of professional translating. But at the same time, some guidance and support are necessary in order not to reproduce the frustration and muddling through caused by the authentic projects.
13The questions I had to answer in order to come up with an adequate method were thus: What abilities do students need to successfully work on authentic projects? And, more importantly, how can we help them acquire these abilities? My solution was to match research findings on translation with a teaching approach based on the constructivist paradigm.
14The reason I chose constructivism as the basis for my instructional endeavour is that it allows for the integration of a multitude of aspects, without requiring some kind of an organised whole as the basis of what is taught. Constructivist theories don’t require the existence of one single, coherent world. Instead, constructivism points to the interpretations each individual makes of his or her sensory input to yield as many worlds as there are people (Thissen 1997). This tallies with the view of translation as a situated activity of an expert nature.
15The characteristic flexibility of expert activity can be triggered by putting constructivism into practice. More exactly, a multi-faceted environment and appropriate tasks can be presented in such a way as to sensitize students to all possible types of translation projects without having to literally teach each of them.
16What the situated paradigm in particular adds to the above considerations is the specific contribution of the context to learning. As an integral part of each individual’s cognitive processes, it can contain factors that hinder learning as well as factors conducive to it, in a different way for different people at different points of the exercise (Risku 2000: 85).
17The challenge in creating a constructivist instructional setting for a situated activity like translation can then be stated as follows: How can a wide variety of factors be presented, and how can students be brought to perceive as many of them as possible as resources? For example, if I would like them to use a computer programme like the CAT tool with hundreds of functions, how do I make students rely on automation and help texts to carry them through? How do I prevent students from becoming overwhelmed by tasks without excessively limiting their own problem-solving? In short: How do I create a setting that comes reasonably close to the complexity of professional translating and thus help students’ transition to this new territory?
18My answer to the challenge outlined in the last section was to create a “translation project management simulation”. Though the simulation itself is by far the most important part of the course, taking up 18 out of 24 academic hours, there is an introduction as well as a very brief summary. The introductory part comprises a synopsis of the research on the translation process, as well as a general introduction to project management. It is intended to give students some background information, which can serve as a basis for reflection during and after the course. Both were presented in a shorter and more to-the-point manner (6 hours down from 12) for the second iteration of the course.
19During the actual simulation, students are first given a rule sheet (presented in Appendix A), telling them what to expect during the simulation and setting the tone for the ensuing game. As explained in the instructions, the instructor, sent them ‘calls for offers’ by e-mail. Each of these comprises a source text as well as a specification of the desired ‘service’ (translation, terminology work and translation, translation and proof-reading or varying numbers and combinations of target languages). This way of giving out tasks had a double purpose. First, it was to make students realise that translation projects also involve non-translation tasks. Secondly, it was to place students in different roles, sometimes as translators or project managers, and sometimes just as proof-readers with lower stakes in the respective project. (See Figure 5 for a schematic illustration of how the simulation game works.)
Figure 5. The scheme of the simulation
20The succession of the ‘projects’ for which students ‘bid’ was arranged by order of increasing complexity. Complexity was defined with reference to the instructional setting by the number of functions of the CAT tool and the number of roles to co-ordinate.
21Successful participation in the ‘bidding’ as well as the ‘working’ phases of the simulation both require students to use the CAT tool, but with varying depth and intensity. For e-mail, a word processor and a spreadsheet programme are equally advantageous.
22If and how much of any of these technologies had to be used was for students to find out through the completion of their tasks. For example, counting words or sentences, and having the number of matches from the translation memory calculated without the tool turned out to be extremely time-consuming. This, in turn, led students to return to the CAT tool’s functions and what they had previously learned about it.
23Students also had to ask for classmates’ co-operation and agreeing on ‘prices’. This task is not covered by the CAT tool. Because of this, informative side-effects occurred when students still tried to use the tool for this task. As nearly every student had created a project artefact for each project during the ‘bidding’ phases, a great number of project icons were displayed on students’ screens. As a result, students wondered when they should ever work on all these ‘projects’. This is just one example where students could infer important information from their (incorrect) use of the CAT tool. The latter thus was not only the subject matter of the course, but also an indirect source for learning within the framework of the simulation. In agreement with the constructivist paradigm, the simulation was thus aimed at helping students “construe knowledge themselves with the help of questions and information” (Thissen 1997).
24One last important issue was to make students see their classmates as resources, as they would see their colleagues in a professional setting. In order to achieve this, questions to me as the instructor were limited to ‘office hours’ of no more than 10 minutes every hour or so. In the meantime, I explicitly referred students to their peers, and slipped completely into the role of a ‘buyer’.
25The objective on which simulation-type teaching was that students were supposed to gain a better insight into professional translating by completing tasks of increasing complexity. Complexity, however, was supposed to be mitigated by way of learning enhancing context factors, as suggested by the situated paradigm. In addition, a certain need for guidance was also to be catered for through the possibility of asking the instructor questions during the time-outs called ‘office hours’. Both these factors were intended to close the gap between simply language-oriented courses and real-world complexity. On a more concrete level, I wanted to see if the situated-constructivist approach helped students to successfully complete the would-be ‘projects’, and if they could use CAT tools adequately for this purpose.
26Two kinds of data were used to test this hypothesis. First, I had students complete a questionnaire. Central questions are paraphrased below, following the numbers in brackets. Out of the 11 questions (12 for the second iteration), 5 (6 for the second iteration) were intended to evaluate the instructional outcome perceived by the students:
27 Could they conceive why the topics treated were being treated?
 Had their learning expectations been met?
 Did they think the course prepared them for their professional life?
 Did the relation between theory and practice become clear?
 Were they incited to look more into using CAT tools?
 Did they get help with individual questions? (used for the second iteration only).
28The rest of the questions were related to the instructor’s teaching style. The questionnaires include a section in which students had the possibility to make any additional free comments on the course.
29The answers to these questions are indicators of whether students could make sense of the course, and relate the teaching material, presented in no easily discernible order, to their studies. To identify hints of success, but also pitfalls of the learning process and method, I complemented the results with students’ questions and comments and my observations of their actions during the simulation. In this way, I was able to develop an idea of how learning really took place, or why it did not, in some instances.
30The results presented hereafter refer to two iterations of the same course, conducted in the winter terms of the academic years 2010/11 (9 questionnaires out of 15 participants) and 2011/12 (4 out of 6) respectively. Although the basic framework of the simulation, as described in section 4, remained unaltered, the introductory phase of the course was shorter for the second iteration. In addition, since group dynamics and the abilities of different participants play a role in a classroom with much student activity, students’ answers to the same question on the questionnaire cannot be interpreted strictly as referring to the same context each time.
31Having said that, the results are comparable for both iterations, at least on the extremes of the spectrum. Thirteen of the 13 students surveyed agree that they were “incited to look more into using CAT tools.”, and 11 out of 13 think “the course prepared them for their professional life”. Ten and 9 respectively agreed that “they [could] conceive why the topics treated were being treated” and that “[t]heir learning expectations [had] been met”. These results show that the general idea of conveying an image of real-world translating through the simulation worked. Yet some doubts remain as to whether students really understood what they were doing and why. But with 4 out of 4 claiming that they “[got] help with individual questions”, this may be more relevant in theory than in practice. In fact, there might be such a thing as a shortcut to success, with students making progress in their practical handling of translation projects and the CAT tool used in completing these projects without consciously referring to any theoretical concept.
32Two observations support this view. The first one, from a certainly strong learner from the second iteration, goes as follows: When, near the end of the course, I demonstrated a function making it possible to split texts so as to divide tasks up between several translators, proof-readers and editors, he asked me whether terminology work would also be split into work packages being assigned to different persons. He went on to argue that this wouldn’t make sense, as terms would get “messed up”. So, without mentioning the term, this student had correctly remembered that terminology work served, among others, to ensure the quality of a translation (consistency being one of its factors), and concluded what entrusting different people with tiny bits of that task could lead to. If recordings had been made during the study, these key findings would surely be mirrored in key quotes, issued from computer-assisted qualitative data analyses (caqdas).
33The second point is a general observation that students are able to perform a task (or a sub-task, a step in using the software) better when it comes up during a simulation ‘project’ the second time. A possible reason for this may be that they learn through repetition rather than transfer.
34On the other hand, a number of problems occurred, most of which can be subsumed under one or several of the following headings:
35 Students have a lot of difficulty with using CAT and other common software in general. This includes them being uncomfortable with a “trial-and-error” approach for exploring new functions, and their generally not reading error messages.
 Students cannot conceive of the CAT tool’s process automation function. For example, they often do not understand the distinction between a change (in the order of steps) being made on the process or document level.
 Students have great difficulty in seeing language artefacts, like terms or sentence pairs, as resources. Seeing a translation memory analysis of a source document, the majority cannot make sense of the data given.
 Students have hardly any idea of what it means to be involved in a for-profit activity sold on a ‘market’.
36To illustrate what this meant in practice, let me give two examples: As a bad surprise, some students, when being e-mailed their ‘calls for offer’, simply did not know how to save the attached document on their computer, let alone name it in such a way as to be able to retrieve it from a folder on a standard personal computer. This problem carries over to working within the CAT tool, as the same principle of document management is used to structure projects in tasks there.
37With respect to the artificial ‘market’ the simulation was meant to create, some students refrained from the ‘project management business’ completely, accepting an occasional translation task. Far from being a simple disciplinary problem, this cut students from an important context factor: By using the tool, they could have seen a visual representation of project work. This might have contributed to their better understanding of the general need for efficiency in general and of automated processes in particular.
38As the results section has shown, there is reason to believe that the translation project management simulation has some positive learning and motivational effects. However, evidence for the proper functioning of the simulation type classroom as such is much less clear. In particular, students hardly ever referred to translating as a “process” or “project” (even implicitly), and most of them seemed to acquire the know-how needed for task – appropriate use of the CAT tool by mere repetition and mimicry. This is undesirable in so far as it may hinder their successfully transferring their learning. For, if the motivation for a constructivist approach was to initiate them to a multitude of project characteristics and their variation, this is supposedly not happening through the type of “shallow” learning that seems to be going on.
39The problem types observed, far from passing judgement on students, contain valuable information on what hinders them from crossing the gap mentioned above between text-centred classwork and settings with a greater or lesser degree of authenticity and complexity. They could thus be turned into routes of improvement for the teaching of translation. This can be arranged around the multitude of functions provided by modern CAT tools while going beyond the situatedness paradigm to include contexts of use and interactive patterns.
40From an instructional standpoint, it also seems very desirable to develop and test a sounder concept of task complexity (or difficulty). If students have a hard time transferring part of what they have learned from one ‘bidding phase’ to the next, this may well be due to the task model not incorporating enough factors relevant to making a smooth transition. Ideally, this model should thus be integrated into the CAT tool (or some specific module thereof) in order for it to function independently from the person of the instructor.
41To conclude, the simulation setting seems to open up new paths for translator training as well as research on translation projects. The setting has also proved to be very flexible in catering for the needs of students with different language backgrounds. It can also be tailored to accommodate a wide range of individual levels of ability in the same course. To conclude, it is a promising endeavour into the future of academic and non-academic translator training.