Essentially these are exams in which students are allowed to take into the exam room reference material such as dictionaries or grammars. Alternatively, or in addition, students might be given FL articles or taped material that they can refer to during the exam. Some open exams may extend over longer periods than the usual two or three hours to allow candidates to work through the material at their own pace.

Academics tend to be divided on such approaches. Arguments against them include:

• Exams and in-class tests are not a true examination of linguistic ability if students have recourse to anything other than their FL knowledge and skills.
• If students are not forced to learn vocabulary and grammatical forms, they will not do so.
• Some students are likely to bring more or better reference materials with them, thus opening the exam up to charges of not being equitable. And it is unlikely to be possible to buy sufficient copies of standard reference works to supply all candidates.
• Space limitations mean it is not always possible to find an exam room with sufficiently sized desks for open-book exams.
• Different standards need to apply if students are to be allowed dictionaries in, say, a translation. This may mean creating additional criteria.

On the other hand:

• Traditional exams are artificial: in professional contexts, language users will have access to all kinds of paper and electronic reference materials; it is unreasonable to expect any learner to demonstrate encyclopaedic knowledge of the FL.
• Open-book exams assess how well students can access and apply information, precisely the transferable information retrieval skills they are likely to need in 'real-world' use of languages.
• They remove the disadvantage that slow writers suffer under in traditional exams.
• They allow more realistic and pedagogically more valuable integrated-skill activities, such as: listen to the tape and write a response, or write an essay based on the information supplied in the accompanying tape and article.
• Translation tests with dictionaries shift the emphasis away from the lexical level to meatier discourse-, function- and culture-related translation issues.

The question is thus not a black and white one. As so often in assessment, a mixed economy taking account of purpose and circumstances is probably the best answer.