Many Japanese learners make mistakes with "a" and "the." In a 2002 study, 80 Japanese university students chose articles and then explained why. Many used "the" when they had a particular thing in mind.

But the real rule is about the listener. Use "the" when the listener also knows which thing you mean. For example, say "I bought a book yesterday." You know the book, but your friend does not.

Japanese has no articles, so it is not surprising that Japanese learners find "a" and "the" difficult. However, a 2002 study by Butler showed that their mistakes follow patterns. She asked 80 Japanese university students to fill in articles and then interviewed them about each choice. Their errors came from different ideas about three points, namely whether the listener knows the thing, whether it is a specific thing, and whether it is countable.

A common mistake was choosing "the" because the speaker had a specific thing in mind, even though the listener did not know it. Speakers of other languages without articles, such as Russian and Korean, show a similar tendency. The good news is that a 2011 study found explicit instruction on articles still had effects six weeks later.

Article errors are among the most persistent problems in Japanese learners' English, and it is tempting to blame the absence of articles in Japanese. Butler's 2002 study, however, revealed that learners operate with their own systematic theories. After 80 Japanese university students completed an article test, she interviewed them about the reason for each choice.

Their errors reflected misunderstandings of three concepts: whether the hearer can identify the referent, whether the referent is specific, and whether the noun is countable, and the pattern of misunderstanding varied by proficiency. A typical error was selecting "the" simply because the speaker had a particular entity in mind, even when it was new to the hearer.

Ionin and colleagues observed a similar overuse of "the" with specific referents among speakers of Russian and Korean, which also lack articles, although whether their account applies equally to Japanese speakers requires separate testing. Encouragingly, Akakura found that computer-based explicit instruction on articles produced effects that lasted six weeks and appeared even on time-pressured tasks intended to measure implicit knowledge, though the validity of such measures is debated. For learners, the key question before writing "the" is simple. Does my listener know which one?