Memory

How do you remember the words you look up while reading?

Look up only the words that matter to the story or keep coming back, save each one with the sentence you found it in, and then test yourself on it: once right after reading, again the next day, and at growing gaps after that. Looking a word up shows you its meaning. Recalling it from memory, several times over weeks, is what makes it stay.

By the Readlang Mobile team · Updated 1 November 2026

Why do you forget a word you have already looked up?

Because looking a word up hands you the meaning without asking you to produce it. Each lookup is one more time you have seen the answer; it is not a time you have recalled it. Memory research separates the two, and recall is what makes a word last. Karpicke and Roediger (2008) had students learn 40 Swahili–English word pairs. Students who kept being tested on every word recalled about 80% a week later. Students whose words were dropped from testing after one correct answer recalled only about a third (roughly 35%), and extra rounds of simply studying those words made no measurable difference.

Even deliberate study fades when it happens in one sitting. Webb, Yanagisawa and Uchihara (2020) pooled 100 effect sizes from 22 studies of four word-focused activities: flashcards, word lists, writing and fill-in-the-blanks. On average, learners could recall the meaning of 60.1% of the words right after the activity and 39.4% on delayed tests; for recalling the word form itself, the figures were 58.5% and 25.1%. Results for individual activities ranged from 18.4% to 77.0% on the immediate tests. Roughly, a third of the meanings and more than half of the forms that learners knew right after a session were gone by the later test.

Reading alone keeps even less. In Waring and Takaki's (2003) study, learners who read a graded reader could translate only a few of 25 new words straight afterwards and, on average, about one of them three months later. A word you have looked up six times and still do not know is the normal result of meeting it six times without once being asked to remember it.

Which words should you look up and save?

Save the words you need to follow the text and the words that keep coming back. Skip the rest, or look them up and let them go.

• Words the story depends on. Peters (2007) gave Flemish university learners of German a text with an online dictionary. Some target words were needed to answer the comprehension questions, others were not. Learners looked the needed words up more often and clicked on them more, and only this relevance affected how many words they remembered; the effect was still there on a delayed test. A word you looked up because you could not follow the sentence without it is a better candidate for a card than one you looked up out of curiosity.
• Words you meet again. Most different words in a book are rare in that book. The table shows how many appear exactly once. A word that has already turned up two or three times is likely to turn up again, which makes it worth the review time; this rule of thumb is our suggestion, not a research figure.
• Frequent words in general. The research on vocabulary size suggests learning the 2,000–3,000 most frequent word families deliberately. If a word you look up is one of them, save it even if this book uses it only once. Our list of the 500 most common English words in books shows which those are for English.

Most of the text is carried by a smaller set. The Great Gatsby uses 5,940 different forms; its 1,926 most frequent forms make up 90% of the text, and 3,164 forms occur only once. The full tables are on how many different words are in a book.

Should you guess from context before looking a word up?

A quick guess is fine, but do not expect the guess itself to make the word stick, and always check it. Mondria (2003) compared two ways of learning the meanings of new words. In one, learners inferred the meaning from context, checked it against a word list and then memorised it. In the other, they were given the translation and memorised it straight away. Both led to similar retention, but inferring took considerably more time. Of the three steps, memorising contributed most; checking contributed about as much as inferring.

Unchecked guesses can also go wrong. In Hulstijn, Hollander and Greidanus's (1996) study, meeting an unknown word again helped advanced learners of French more when they were given its meaning or looked it up than when they read without help. The authors' explanation: readers left on their own often ignore unknown words or guess them wrongly.

Guessing works best in easy text. At about 98% known words, one unknown word in fifty, the surrounding sentence is usually clear; at 90%, it often is not. Where these figures come from is explained under coverage.

What should you save with each word?

• The sentence you found it in. It reminds you of the meaning that fits, which a dictionary entry with ten senses does not, and it can become a card with the word blanked out.
• The meaning in that sentence, in your own language or in simple words of the language you are learning.
• The dictionary form as well as the form you met. In languages with many endings, a book shows the same word in many spellings. A note such as hablaban (from hablar) saves you making separate cards for each form.

Start with cards that show the foreign word and ask for the meaning. Add cards in the other direction only for words you want to use when speaking or writing, as the practical settings for spaced repetition suggest.

Does writing your own sentence with the word help?

Not as much as it seems, and early on it can get in the way. Barcroft (2004) taught learners of Spanish 24 new words from word–picture pairs in two experiments. When learners wrote each word into a Spanish sentence, they did worse on immediate and delayed tests of producing the words than when they only studied the pairs, a strong negative effect in both experiments. Barcroft concluded that sentence writing can hinder learning the word form in the first stages of learning a word.

Folse (2006) compared three exercises with 154 university students of English: one fill-in-the-blank exercise, three fill-in-the-blank exercises, and one exercise of writing an original sentence. Words practised in three fill-in-the-blank exercises were retained much better than words from the other two conditions. His conclusion was that what counts is the number of times a learner has to retrieve the word, not how deeply the word is processed once.

So spend the time on recall instead. Three short tests of a word over the next days do more than one careful sentence today. Writing sentences is still good practice for using words you already know.

How should you review the words after reading?

Do a short recall round right after you finish reading, then review the saved words at growing gaps.

A task after reading helps by itself. Peters, Hulstijn, Sercu and Lutjeharms (2009) had 137 college students in Belgium read a German text with 16 target words. A vocabulary task after reading substantially improved how many meanings they remembered, on both immediate and delayed tests. Announcing a vocabulary test before reading made students look up more words, but meaning retention improved only with the task after reading and for words the comprehension questions made relevant.

After that, spacing does the work. In Cepeda and colleagues' (2008) study of more than 1,350 people learning facts, the best gap was about a day when the final test came a week later, and about three weeks when it came 70 or 350 days later. The research is summed up on the spaced repetition page. A schedule that follows it:

Each time, try to produce the meaning before you turn the card, and count it as missed if you could not, however familiar the answer looks afterwards. That is the step a lookup never gives you.

What if a word still will not stick?

Some words fail again and again. Anki calls them leeches: by default, when a card has been forgotten 8 times, Anki tags it as a leech and suspends it (Anki manual). The manual's advice applies to any method:

• Change the card. The manual calls changing how the information is presented the most efficient fix, for example by adding a memory aid or cards for related words. For a word from reading, a second example sentence from another text is an easy change.
• Separate words you confuse. If two similar words keep swapping, learn one at a time and bring the second back once the first is secure.
• Let some go. A rare word that costs a minute a week may not be worth it. You can drop the card and meet the word again in reading later.

Do you still need to meet the word in books?

Yes. Cards give a word recall; reading gives it more meetings in new sentences, which build what a card cannot: how the word is used, what it goes with, how it sounds in a sentence. Readers of an unsimplified novel recognised the meaning of 84% of the new words they met ten or more times, and one encounter or a few did much less (Pellicer-Sánchez & Schmitt, 2010). Reviewing a word on cards and then meeting it again in a book is what moves it from known on a card to known in reading. The page on learning words from reading covers how many encounters a word needs.

Notebook, Anki or a reading app: how do the options compare?

All of them can work. The question is how much effort saving takes and whether review is scheduled for you. Choose by where you read and by whether you want the review to happen in the same tool.

Readlang Mobile is made by the team that writes this site and is not affiliated with Readlang (readlang.com). If scheduled review matters most to you, Anki next to any reader, Readlang on the web or LingQ does that. The reading tools comparison covers how each tool handles lookups while reading.

A paper notebook makes you choose, which is useful: writing a word down by hand is a small cost that stops you saving everything. Tools that save every lookup automatically are faster, but they fill up with one-off words. Whichever you use, prune the list to the words that matter and recall them on a schedule.

References

1. Anki Manual. Leeches. docs.ankiweb.net, checked 31 October 2026.
2. Barcroft, J. (2004). Effects of sentence writing in second language lexical acquisition. Second Language Research, 20(4), 303–334. doi:10.1191/0267658304sr233oa
3. Cepeda, N. J., Vul, E., Rohrer, D., Wixted, J. T., & Pashler, H. (2008). Spacing effects in learning: A temporal ridgeline of optimal retention. Psychological Science, 19(11), 1095–1102. doi:10.1111/j.1467-9280.2008.02209.x
4. Folse, K. S. (2006). The effect of type of written exercise on L2 vocabulary retention. TESOL Quarterly, 40(2), 273–293. doi:10.2307/40264523
5. Hulstijn, J. H., Hollander, M., & Greidanus, T. (1996). Incidental vocabulary learning by advanced foreign language students: The influence of marginal glosses, dictionary use, and reoccurrence of unknown words. The Modern Language Journal, 80(3), 327–339. doi:10.1111/j.1540-4781.1996.tb01614.x
6. Karpicke, J. D., & Roediger, H. L. (2008). The critical importance of retrieval for learning. Science, 319(5865), 966–968. doi:10.1126/science.1152408
7. Leitner, S. (1972). So lernt man lernen. Freiburg im Breisgau: Herder.
8. Mondria, J.-A. (2003). The effects of inferring, verifying, and memorizing on the retention of L2 word meanings: An experimental comparison of the “meaning-inferred method” and the “meaning-given method”. Studies in Second Language Acquisition, 25(4), 473–499. doi:10.1017/S0272263103000202
9. Pellicer-Sánchez, A., & Schmitt, N. (2010). Incidental vocabulary acquisition from an authentic novel: Do Things Fall Apart? Reading in a Foreign Language, 22(1), 31–55. Full text (ERIC)
10. Peters, E. (2007). Manipulating L2 learners' online dictionary use and its effect on L2 word retention. Language Learning & Technology, 11(2), 36–58. hdl:10125/44103
11. Peters, E., Hulstijn, J. H., Sercu, L., & Lutjeharms, M. (2009). Learning L2 German vocabulary through reading: The effect of three enhancement techniques compared. Language Learning, 59(1), 113–151. doi:10.1111/j.1467-9922.2009.00502.x
12. Waring, R., & Takaki, M. (2003). At what rate do learners learn and retain new vocabulary from reading a graded reader? Reading in a Foreign Language, 15(2), 130–163. ERIC record
13. Webb, S., Yanagisawa, A., & Uchihara, T. (2020). How effective are intentional vocabulary-learning activities? A meta-analysis. The Modern Language Journal, 104(4), 715–738. doi:10.1111/modl.12671
14. Readlang. Pricing. readlang.com, checked 31 October 2026.
15. Apple App Store (US). Readlang Mobile (listed as Readlang Language Readers). apps.apple.com, checked 31 October 2026.
16. Readlang Mobile. Home page (Premium features). readlang-mobile.com, checked 31 October 2026.
17. Apple App Store (US). AnkiMobile Flashcards, Anki Software LLC. apps.apple.com, price checked 31 October 2026.