Headword Band
The controlled-vocabulary specification a graded-reader author writes to. Each band names a maximum number of headwords permitted in the text (typically 250, 400, 600, 800, 1200, 1800, 2500, 3000 in the major series), where a headword is a word family covering inflections and common derivations. The band is the primary authoring constraint: the writer can use any item within the band freely, must use items beyond it sparingly and with support, and is expected to recycle the band's higher-frequency items densely enough that learners meet them many times per book.
What the band controls
A band is the writer's working vocabulary plus a small margin of items learners can recover from context, glossing, or footnotes. The working vocabulary covers the words the writer may use without restriction; learners at this level are assumed to know them. The margin covers items above the band that the writer judges essential for plot, setting, or topic: typically proper nouns, a small number of low-frequency content words, and culturally embedded terms. Claridge (2012) found that publishers vary in how strictly they police that margin. Cambridge English Readers permit no glossary at all, forcing all in-band recovery; Oxford Bookworms supplies running footnotes for cultural items; Penguin and Macmillan provide end-glossaries. The band itself is treated as a hard ceiling on the working vocabulary across all surveyed series.
Publisher band ladders
The major series anchor their bands to internal frequency lists derived (with adjustments) from sources such as the BNC, COCA, and the General Service List family. Hill's ELT Journal surveys (1997, 2001, 2008, 2013) document the resulting ladders.
- Oxford Bookworms Library: Starter (250), Stage 1 (400), Stage 2 (700), Stage 3 (1000), Stage 4 (1400), Stage 5 (1800), Stage 6 (2500).
- Penguin Readers (Pearson): roughly 200, 300, 600, 1200, 1700, 2300, 3000 across seven levels.
- Cambridge English Readers: 250 to 3800 across seven levels, the broadest range of the major series.
- Macmillan Readers: 300 to 2200 across six levels, with a strong non-fiction strand.
Cross-series alignment is approximate. Claridge's interview study showed that a 1400-headword Oxford text and a 1400-headword Macmillan text are not equivalent in difficulty: the underlying lists differ in coverage of low-frequency items, and the grammar syllabuses paired with each band differ. CEFR mappings published by the series are best treated as orientation rather than equivalence.
Relationship to the BNC/COCA frequency lists
Most current band lists derive, directly or by adjustment, from corpus frequency. Nation and Waring's (2020) treatment positions the BNC/COCA word-family lists as the most defensible base for graded-reader construction: bands at 1000-family intervals up to 9000 give the writer empirically grounded targets, and any text profiled against these lists yields a clear coverage report. Several publishers have moved partly toward such lists, but most retain proprietary lists tuned for pedagogical rather than corpus-frequency considerations (e.g., promoting a slightly less frequent word into a low band because it is high-utility for learners). The headword band as a concept is corpus-grounded; specific publisher bands are corpus-informed at best.
Authoring constraint, not a label
The band is felt most concretely at the desk. Writing a Stage 2 (700-headword) text means rejecting replied in favour of answered, enormous in favour of very big, suddenly in favour of and then. The discipline is harder than it appears, because high-frequency English carries less semantic specificity, and the writer must compensate through plot, dialogue, and structure. Hill (2008) repeatedly notes that the most successful graded readers are those whose authors treat the band as a creative constraint rather than a vocabulary cap.
References
- Claridge, G. (2012). Graded readers: How the publishers make the grade. Reading in a Foreign Language, 24(1), 106–119.
- Hill, D. R. (2008). Graded readers in English. ELT Journal, 62(2), 184–204.
- Hill, D. R. (2013). Graded readers. ELT Journal, 67(1), 85–125.
- Nation, I. S. P. & Waring, R. (2020). Teaching Extensive Reading in Another Language. Routledge.