Keeping a language you already have

The experience of returning to a language you once spoke well and finding it gone is common, and it is not a failure of memory so much as a predictable consequence of disuse.

What attrition looks like

The pattern.

Vocabulary retrieval slows first, particularly for less common words.

Production degrades faster than comprehension, which is why people can still follow a conversation they cannot participate in.

Fluency and confidence decline before accuracy.

Accent and basic grammar are relatively resistant.

And the loss is largely of access rather than of storage, which is why a period of re-immersion restores things faster than the original learning took.

How much use is enough

The practical question.

Considerably less than learning required.

A few hours a week of input and some production appears to hold most people steady, though this varies with level.

Frequency matters more than duration: twenty minutes daily beats two hours on Sunday.

And the level reached matters: a language taken to a high level is substantially more resistant to loss than one abandoned at intermediate.

Input for maintenance

The efficient half.

Podcasts and radio during commutes and chores, which costs no additional time.

Television and film in the language.

Following news sources in the language rather than in your own.

Reading, which maintains vocabulary better than listening.

And switching devices or accounts to the language, which produces constant small exposure.

Input alone maintains comprehension well and production poorly, which is why it is not sufficient.

Production for maintenance

The half that is skipped.

Regular conversation — a tutor, an exchange partner, a friend — even monthly.

Writing: a journal, messages, correspondence.

Speaking to yourself, describing what you are doing.

Thinking in the language deliberately during routine activities.

And translating something occasionally, which forces active retrieval of specific words.

Heritage languages

A specific and common case.

A language spoken at home in childhood and not maintained frequently ends up as strong comprehension and weak production, with a vocabulary limited to domestic topics.

Which is a distinctive profile rather than an incomplete version of an adult speaker's language.

What helps: formal study to add the registers never acquired at home; reading, which builds the vocabulary childhood conversation did not; and speaking with adults about adult subjects.

And the emotional dimension is real: for many people this language is tied to family and to identity, which makes both the loss and the recovery weightier than a language learned in a classroom.

Multiple languages

Where the competition is.

Maintaining several requires deliberate rotation, since the strongest and most-used will crowd out the others.

Which suggests: assign each a different activity or context, so they are not competing for the same slot.

One for reading, one for podcasts, one for a monthly conversation.

And accept that the weakest will decline, and decide deliberately whether that is acceptable rather than discovering it later.

Interference

A known irritation.

Learning a new language frequently disrupts access to a previously learned one, particularly if they are related.

Words from the wrong language surface, which is a retrieval issue rather than a loss.

It generally resolves as the new language stabilises.

And deliberate separate practice for the older language during the period of learning the new one reduces it.

Recovering a lapsed language

The good news.

Relearning is markedly faster than learning, and there is evidence that even languages apparently entirely lost from childhood leave traces that accelerate reacquisition.

Which means the return is worth attempting even after decades.

Practical route: heavy input first, which reactivates comprehension quickly, then speaking practice, which is where the discomfort is.

And expecting a period where comprehension has returned and production has not, which is normal rather than a plateau.

Making it sustainable

The general principle.

Maintenance that depends on discipline fails; maintenance embedded in something you would do anyway does not.

Which means the goal is to find a genuine use — a community, a hobby, a set of friends, a category of reading you prefer in that language.

A language attached to something you want is maintained without effort. A language attached only to the ambition of keeping it is not.