A 32-year-old sales analyst in Taoyuan sees allocate on an Anki card and chooses 「分配」within 2 seconds, yet during Friday's budget meeting the phrase “allocate the remaining funds” never arrives. Recognition gave the learner access from English to meaning. Speaking required the return route from an intended meaning to an English form. Vocabulary cards should train the direction the learner expects to use.
This article is for an intermediate adult who recognises thousands of words yet uses far fewer in meetings and presentations.
Knowing a word has directions
Receptive knowledge supports recognising a word in reading or listening. Productive knowledge supports retrieving and shaping it for speech or writing. The boundary is gradual. A learner may recognise allocate, recall its core meaning, forget which preposition follows it, and still understand a report. Producing “allocate 17 percent of the budget to training” asks for spelling or sound, grammar, collocation and meaning at once. One card direction samples one route. English on the front and Chinese on the back mainly asks for receptive meaning. Chinese or a situation on the front, with English recalled before turning the card, asks for productive form. A multiple-choice app makes recognition easier again because the answer remains visible among the options.
Choc Education's article on the mental lexicon and word games describes words as connected entries rather than a flat list. Retrieval direction is one property of that network. A link used only from form to meaning may remain slow in reverse.
Vocabulary size changes with the test
Webb examined receptive and productive vocabulary size with equivalent translation formats and separate target words. Learners' total receptive vocabulary was larger than their productive vocabulary. Under stricter scoring for fuller knowledge, the receptive advantage appeared in all three frequency bands and grew for less frequent words. Under partial-credit scoring, the difference within each band was small. The 2008 Studies in Second Language Acquisition paper shows why a single vocabulary number can mislead. A prompt, scoring rule and direction define what “known” means on that test. The finding is descriptive. It does not prove that reversing every card will close an individual's gap. Translation tests also capture only part of word knowledge and cannot reproduce time pressure, turn-taking, pronunciation demands or interruptions during a fast client call. They do establish a practical diagnostic: test the same target in the direction the learner needs.
Learning direction leaves different traces
In a 2009 experiment, Japanese learners of English studied word pairs either receptively or productively, and Webb then tested five aspects of knowledge in both directions: spelling, association, syntax, grammatical functions, and meaning and form. That produced 10 tests for each target. Productive learning led to larger gains in receptive and productive spelling, along with productive meaning, syntax and grammatical-function knowledge. Receptive learning produced larger gains on receptive meaning. The RELC Journal study found that the study direction changed both the amount and type of knowledge gained.
Receptive cards remain useful.
A learner preparing to read 26 pages of an engineering manual needs fast recognition during a 12-minute pre-reading review, and producing every technical label may spend time poorly. Beginners can also freeze when a broad Chinese cue admits six possible English answers. Productive prompts work best when the expected answer is constrained by a phrase, sentence or image. I would rather keep a well-designed receptive card than reverse a vague pair such as 「處理」and handle. The Chinese cue could call for deal with, process, manage, address or handle. A sentence fragment like “___ a customer complaint” narrows the decision and teaches a usable partnership.
Give high-use words two routes
Select words by future use. From a list of 47 new items, perhaps 11 deserve active command for next week's presentation. The rest may remain recognition targets until they recur. This prevents productive practice from turning every noun in a reading passage into a speaking obligation.
Build two cards for each active item:
| Front | Required response | Back |
|---|---|---|
| allocate | Give a short Chinese meaning and one noun partner | 分配, allocate funds |
| A chart labelled 「把預算的17%分配給培訓」 | Give one full instruction | Allocate 17% of the budget to training. |
Two prompts test different access routes and expose different failures. If funds appears only after the first letter is shown, collocation retrieval needs work. If the full clause omits to, grammar needs attention. If pronunciation makes funds lose its final cluster, the card can carry a short recording. Do not place all clues on the front. A card displaying the English word, Chinese meaning, picture and first letter before the learner responds measures almost nothing. Keep one clear cue, attempt the answer aloud, then reveal the back. Mark “again” when the word required a hidden hint, even if the final response sounded correct.
Retrieval needs a use test
Card accuracy can rise because the order, screen colour or neighbouring cards become familiar, so after every 11 cards, close the app for a 90-second test and explain one decision using three target items. For the budget set, answer: “Which costs would you reduce, and where would you allocate the money?” Record once without notes.
Listen for meaning first.
Check whether the response communicates a defensible allocation, then inspect the target forms. A correct card answer that never appears in the recording remains receptive or cue-dependent knowledge. A near-synonym may solve the communication perfectly. Record that success and decide whether the original target still deserves practice. Delay the second test by 71 hours and change the numbers from 17 percent to 23 percent. Changing one detail blocks memorisation of the complete sentence while preserving its grammatical frame. If the learner retrieves allocate and loses funds, split the pair. If all three words arrive slowly, reduce the active set from 11 to 7 for the next round.
Keep recognition and production connected
Reading, listening, speaking and writing feed one another, so the two directions should meet inside meaningful input. After productive card practice, find the item in a report, podcast transcript or meeting agenda. Notice who allocates what to whom. Later, use the same pattern to describe a new decision. The context supplies grammar and collocation that a bare translation pair hides. Some words can stay receptive for years. A learner may need to understand amortisation in a finance document and never say it. Active practice belongs to words with a foreseeable job. That job determines the cue.
Before Friday, reverse 11 high-use cards, replace vague Chinese cues with phrases, then record a 90-second budget answer with Anki closed.