How do you measure ability in a language? An eight-year-old can speak his native language fluently, but he is not likely to run a business meeting well. He probably has no clue what an "agenda" is, or what it means to get new leads in the pipeline and update them in the CRM.
The person who has studied flashcards for years may have a huge vocabulary, but no ability to string words together coherently because of a lack of structure and grammar.
How are these abilities tested? The eight-year-old may appear fluent and test poorly in a standardised language assessment exam, while the practiced adult learner may score higher on the same exam. Is this an accurate reflection of ability in the language?
Testing language ability is difficult, and though some standards for language assessment exist, most agree to their imperfection. Part of the problem is that testing is built around language usage and the "four skills" (listening, speaking, reading, writing), rather than around the language itself. The child will have the "language" and understand its core patterns intuitively, but not have knowledge of different situations or constructions. The adult may have experience from his own language that allows him to test well in understanding complex constructions and situations, but may have to fudge the system that holds it all together.
It would be nice to be able to separate testing of the core patterns of the language from testing of the various uses of the language and their accompanying vocabulary. As the patterns must be represented through actual examples, however, this is no mean feat.
Testing will remain a difficult and imperfect aspect of language learning. It is important to see it as a necessary tool for representing and monitoring the progress and relative ability of students as they work toward mastery of a foreign tongue.