Translation into L1 has for years been a popular method of testing language learners in higher education. Many students and teachers, however, express their dissatisfaction with it as a testing technique. Newson (1998: 64-65) discusses the main disadvantages of using translation as a test of language competence and highlights the fact that 'it presents the examinee with random translation problems'. The main problems are that texts chosen are unpredictable, and it is impossible for students to prepare systematically because the whole language is being tested.

In order for translation to become a more effective assessment method, both the use of translation in the classroom and the way it is assessed need to be closely linked to the learning outcomes of the particular programme/course. We need to have clear answers to the following questions:

• What skills is the translation testing?
• Do students know what they are supposed to do and why?
• Have they been prepared for the task? Can they prepare for it?
• Are the assessment criteria clear?

As we saw earlier, we need to use a wide range of language skills in order to produce an acceptable translation. An assessment is therefore required that tests students' use of all these skills. For this reason, Malmkjaer (1998) suggests that assessment should involve translating a text on which some initial groundwork has been done in advance of the test, and that assessment should take account of all aspects of the process.

Mason (1987) stresses the importance of seminar-type discussion when evaluating translations. According to this view, we should not only point out errors when marking translations, but should also subsequently discuss them with students. Peer assessment is also mentioned as a useful assessment method, which relieves some of the teachers' load and involves the students more in the task. (See Module 13, section 13.3.2, on peer assessment.)

In order to assess a translation activity satisfactorily, we need to establish precise criteria that will help us make judgements with a degree of confidence. Of course, the question of objectivity in assessing translations is problematic, as all translation is interpretation and is therefore subjective. However, Mason (1987) points out the need to get rid of vague terms such as 'reads well', 'sounds right', 'captures the spirit of …'. These terms should be substituted by more precise and 'objective' terminology. (See Module 14, section 14.5.2, for an example of one such attempt.)

Here is a summary of the criteria that can be used to assess a translation:

• understanding of the source text;
• accuracy of the translation;
• acceptability of the translation (range of vocabulary, consistency, coherence, register membership, etc);
• comparative skills, awareness, problem solving;
• research skills;
• creativity (particularly in translation of literary and media texts);
• critical skills (if using commentary).

The principal task for the marker of a translation should be to check whether the TT does justice to the ST, whether it fulfils the requirements of the brief, and whether it is an acceptable text in terms of the conventions of the target language and culture.

(See Module 14, section 14.5.2, for further reflections on the assessment of translation.)