Vocabulary maintenance for polyglots: one workflow across multiple languages

If you've reached B2+ in three or more languages, you've hit the polyglot's paradox: acquisition was the easy part; maintenance is the hard part. Forget Spanish for six months while you push German to C1, and the Spanish vocabulary you spent years acquiring starts leaking out at an alarming rate.

The good news: maintenance scales much better than acquisition if you have the right system. The bad news: most flashcard tools and language apps were designed for monoglot beginners, not polyglots managing five language pairs in parallel.

This article is for the polyglot audience: how to build a single workflow that maintains vocabulary across multiple languages without burning out, while still allowing for one "active" language at a time.

The polyglot's actual problem

Common framings of polyglot maintenance miss the real bottleneck. The bottleneck isn't:

"Finding time" — 15 min/day per language is enough for maintenance.
"Choosing what to study" — high-frequency content is the answer for any language.
"Motivation" — most polyglots have no motivation problem.

The real bottleneck is switching cost — the friction of jumping between languages within a study session, between tools per language, between deck systems, and between review schedules. Each switch costs working memory and willpower. Five languages × constant switching = exhaustion.

The system below minimises switching cost.

Core principles

Principle 1 — One tool, multiple decks per language

Don't use Anki for Spanish, Memrise for German, and Duolingo for Japanese. The cognitive overhead of three different UIs, three different review systems, and three different progress views is brutal. Pick one tool that supports multiple decks and stick to it.

For most polyglots, that means Mnemo or Anki, both of which let you create per-language decks within one account and review them in one session.

Principle 2 — Capture in the language's natural reading context

Vocabulary acquired in real reading sticks far better than vocabulary acquired from a frequency list. This is true for any language.

Read native content in each language at least twice a week:

News (Le Monde for French, Der Spiegel for German, NHK for Japanese, etc.)
Substack/Newsletter writers in each language
Podcasts with transcripts (huge resource — transcripts are searchable, listening is realistic)
Books on topics you actually care about

Use a Chrome extension (like Mnemo's) that handles any language pair to capture vocabulary at the moment of encounter, in context. The cross-language flexibility matters — you don't want to set up separate capture flows per language.

Principle 3 — One "active" language, others in maintenance

Trying to be active in 5 languages at once doesn't work. Pick one language to push forward at any time (the active language) and keep the others in maintenance mode. Rotate every 6-12 weeks.

Active language: 30-45 min/day. New input + review + output (writing/speaking).
Maintenance languages: 5-10 min/day each. Review only, occasional input.

Total: ~30-45 min for active + ~20-40 min total for maintenance = manageable in a daily 60-90 min language session.

Principle 4 — Rotate priority, don't redistribute

When you switch which language is "active", don't redistribute time evenly. The active language gets the bulk of attention because that's where the cognitive lift is. Maintenance languages stay light and consistent.

Rotation example over a year:

Q1: Active = German. Maintenance = Spanish, Japanese, French.
Q2: Active = Japanese. Maintenance = Spanish, French, German.
Q3: Active = French. Maintenance = Spanish, German, Japanese.
Q4: Active = Spanish (refresh and push to next level). Maintenance = others.

Each language gets one "active quarter" per year — enough to make real progress, not so much that maintenance languages decay.

The daily routine

A polyglot's actual daily session, ~75 min:

07:00-07:25  Active language input (read 10-15 pages of native content,
             capture ~15-20 new words via extension)

07:25-07:45  Active language review (FSRS-scheduled cards from active deck)

07:45-08:00  Maintenance languages review (FSRS surfaces due cards from
             all maintenance decks, mixed)

08:00-08:15  Active language output (write 200 words, or 5 min spoken
             practice with a tutor / yourself)

Notes on this:

Maintenance review is mixed. Don't review Spanish for 5 min then German for 5 min — that's high switching cost. Let the SRS algorithm pull due cards from all maintenance decks into one combined queue. Each card is in its target language, so context-switching happens naturally.
Output only for the active language. Maintenance languages don't get explicit output time. Output for them happens organically through occasional native-content reading and the rare conversation.
Sundays: curate the week's captured words across all languages (~15 min). Add collocations, write personal-relevance sentences, delete junk.

Per-language deck structure

For each language, maintain three deck types:

Core vocabulary — high-frequency words from your acquisition phase. Review in maintenance mode forever.
Active reading — words captured in the current quarter. Higher new-card rate; gets pruned/promoted to Core after 6 months.
Domain vocabulary — field-specific (your work, hobbies, niche reading). Lower review priority but kept accessible.

When a language switches from active to maintenance, the "Active reading" deck stops receiving new cards but continues to be reviewed.

Common polyglot mistakes

Trying to keep all languages "active"

Result: nothing actually progresses. Cognitive load is too high. You feel busy but plateau in all of them.

Mixing languages within a single deck

"I'll have one big polyglot deck!" — sounds elegant, fails in practice. The card type, the example sentence language, the audio language, the translation direction — they all differ per language. Mixing creates confusion and wrong-language retrieval.

Each language gets its own deck.

Skipping maintenance for the active language

Counterintuitive but real: when you go all-in on German for a quarter, you must not skip your Spanish/Japanese maintenance. Otherwise the "switching cost" of returning to Spanish next quarter is enormous. Maintenance is cheap but mandatory.

Using consumer language apps as the primary tool

Duolingo, Memrise, Babbel — these are designed for beginners who need motivation and structure. Polyglots past A2 in a language don't need gamified beginner content; they need a high-quality SRS that handles native-context vocabulary.

The dedicated tools — Anki, Mnemo, RemNote — outperform consumer apps for any learner past A2.

Tooling specifics for polyglots

What you need

Multi-deck support within one account.
FSRS (or comparable modern SRS) — for memory efficiency, especially important when reviewing 4-5 languages.
Cross-language capture — a Chrome extension or similar that doesn't care what language is on the page.
Combined review queue option — for the maintenance-mix step in the daily routine.
Sync across devices — you'll review on phone, web, and desktop in the same week.

Tools that meet these requirements

Mnemo — multi-deck, FSRS v5 default, Chrome extension supports any source/target language pair, syncs across web + extension + macOS. Free.
Anki + FSRS — most flexible, largest deck-sharing ecosystem, supports custom card types per language. Steeper learning curve. Free on desktop, $24.99 iOS one-time.
RemNote — FSRS in Pro tier, good if you also use it for note-taking.

For most polyglots starting fresh, Mnemo is the lowest-friction starting point. For polyglots who've already invested years into Anki workflows, stick with Anki and enable FSRS.

A note on dictionary quality

Polyglots often hit limits with the built-in dictionaries of consumer tools. For serious work:

Wiktionary — best free multilingual dictionary; covers etymology, conjugation, IPA across most languages.
Linguee — bilingual examples sourced from real translations; excellent for verifying collocations.
Native-language monolingual dictionaries at C1+ level (Le Robert for French, Duden for German, etc.) — using monolingual definitions strengthens your in-language thinking.

A good Chrome extension should let you switch between dictionary sources or fall back gracefully across them. Mnemo's extension uses dictionary APIs across many languages with translation as the fallback.

Summary

Polyglot maintenance is solved by reducing switching cost:

One tool, multiple decks — same UI, same review system across all languages.
Capture in real reading context for each language, with a tool that handles any language pair.
One "active" language at a time, others in light maintenance, rotate quarterly.
Daily routine — 60-75 min covering active input + review + maintenance + output.
Avoid consumer apps past A2 — they're designed for beginners.

Try Mnemo for this workflow — free, supports any language pair, FSRS v5 by default, multi-device sync.