Key evidence an nummers
| Coverage o kent wirds | Unkent tokens in a 300-wird passage | Meanin |
|---|---|---|
| 90% | 30 | Example o the arithmetic |
| 95% | 15 | Aboot ae unkent token in 20 |
| 98% | 6 | Aboot ae unkent token in 50 |
The table is arithmetic aboot text tokens, no a comprehension test. A token is a time a wird appears in the passage: repeat appearances o ae unfamiliar wird are coontit ilka time. Wird faimlies are anither unit uised tae describe wird buik size. Mixin up thae units maks an apparently precise requirement misleadin. Nation, 2006 Brysbaert et al., 2016
Whaur dis the 8,000–9,000 estimate come fae?
Nation uised wird-family leets based on frequency tae examine coverage in written an spoken Inglis. The conclusion is conditional: gin aboot 98% coverage is needit for comprehension withoot help, the written material he examined needs approximately that range o wird ken. Proper names and hou related forms are dealt wi are pairt o the analysis, no details tae leave oot o a universal learnin target. Nation, 2006
Ither research on coontin wirds shaws that lemmas, faimlies an written forms can gie vera different totals. A result o 8,000 faimlies cannae be compared direct wi a test reportin 8,000 lemmas, an it isnae a coont o aw the wirds in Inglis. Brysbaert et al., 2016 Nagy et al., 1984
Is 98% a magic threshold for comprehension?
Schmitt, Jiang an Grabe testit 661 pairticipants fae aicht countries on twa texts. They fand a fairly straicht relationship atween vocabulary coverage an comprehension, withoot a sharp pynt whaur unnerstandin suddenly became complete. Their results back 98% as a useful target for coverage o academic texts, no an aw-or-nae gate. SCHMITT et al., 2011
Whit affects it?
Whyever buiks differ
A story ye ken, a specialist textbook an a historical novel hae different wirds an assume different ken o the subject. The coverage estimate disnae tak grammar, inference or subject ken awa fae comprehension. Research on readin speed likewise shaws why gettin through wirds an unnerstandin them are sindry outcomes. Nation, 2006 Brysbaert, 2019
Receptive ken is the richt stertin pynt for recognisin text, but a learner can ken a wird pairtly. Studies that distinguish recognisin, recall an uisin wirds yersel shaws why markin a wird as familiar disnae guarantee ye can mak full sense o aw its meanings in a passage. Laufer et al., 2004 Webb, 2008
Can ye stert readin wi fewer wirds?
Aye: the estimate is aboot coverage withoot help in the material examined, no permission tae read. Easier material, support an different readin goals change things. A learner uisin a dictionar is daein a different task fae the condition wi nae help. The nummer should help ye choose material, no put ye aff readin till a test reaches a particular total. Nation, 2006
Meetin a wird again can develop different pairts o wird ken, but research on repetition disnae gie a universal nummer o encounters that maks ilka wird mastered. Keep track o comprehension an the kynd o help ye need, insteid o ettlin that a fixed nummer o newly recognist wirds guarantees fluent readin. Webb, 2007
Whit aboot age, sex an ither leids?
This is an analysis o text coverage in Inglis, no a norm for ability based on age or sex. Ither leids coont forms an faimlies in different wyes, an different genres need different wird ken. Applyin the nummer across leids withoot checkin the units would mak the evidence seem stronger than it is. Nation, 2006 Brysbaert et al., 2016
The useful practical question is hou muckle o the buik ye unnerstaun at the level o help ye've chosen. A recognisin score can gie context, but a short comprehension check on the actual material is nearer that goal than a universal cutoff for wird ken. SCHMITT et al., 2011 Laufer et al., 2004
Gie the related free test a try
Wird buik test
Try the wird buik test tae explore recognisin Inglis wirds, then think aboot the buik itsel an yer comprehension. The score isnae a certificate that ilka text will be accessible, and the unit it coonts maun match ony published comparison.
Common questions
Are 8,000 faimlies the same as 8,000 wirds?
No in ilka coontin system. A faimly groups related forms, while a lemma or written form is a different unit. Brysbaert et al., 2016
Dis 98% coverage mean 98% comprehension?
Naw. Coverage coonts kent occurrences o wirds; comprehension means unnerstandin the text. The twa are no the same thing. SCHMITT et al., 2011
Should I wait till I ken that mony?
The analysis is aboot coverage withoot help in particular material. It isnae a rule bannin readin wi support or easier buiks below that range. Nation, 2006
Sources
- Brysbaert et al. (2016). Hou Mony Wirds Dae We Ken? Practical Estimates o Wird-buik Size Dependin on the Definition o a Wird, the Muckle o the Leid Input an the Pairticipant's Age.
- Nation (2006). Hou Muckle Wird Ken Is Needit for Readin an Listenin?.
- Laufer et al. (2004). Testin Wird Ken: Size, Strength an Computer Adaptiveness.
- Webb (2008). Receptive an Productive Wird-buik Sizes o L2 Learners.
- Webb (2007). The Effects o Repetition on Wird Ken.
- Nagy et al. (1984). Hou Mony Wirds Are Thare in Printed School Inglis?.
- Brysbaert (2019). Hou Mony Wirds Dae We Read per Minute? A Review an Meta-analysis o Readin Speed.
- SCHMITT et al. (2011). The Percentage o Wirds Kent in a Text an Readin Comprehension.
Cite this page: Hou mony wirds dae ye need tae ken tae read a buik in Inglis?. Helibrain. 2026. https://helibrain.com/sco/answers/how-many-words-to-read-a-book. CC BY 4.0.
← Explore mair answers