Examining the word family through word lists

Abstract

The choice of lexical unit has important consequences for L2 vocabulary research, testing and instruction. In recent years, the most widely used lexical unit has been the word family. This study examines the characteristics of word lists based on the word family and explores the levels of text coverage such lists may provide should the assumption that learners can deal with word families be incorrect. This is pursued through the detailed examination of a set of word-family-based word lists. The study finds that such word lists pose a number of challenges, including the number of word forms with multiple affixes, the number of word forms with more challenging affixes, and the number of word families in which the base word is not the most frequently occurring member. Moreover, the first thousand word families in particular are shown to be challenging. The study then demonstrates that if learners are unable to deal with the complexity of word families, even to a relatively small degree, word-family-based lists may provide far lower text coverage levels than may be assumed. It concludes that in work on second language vocabulary, careful consideration is needed of the appropriacy of the word family as the lexical unit and highlights the range of work based on the word family that may need reevaluating.

Evaluating lists of high-frequency words: Teachers’ and learners’ perspectives

“…Headwords were chosen as the unit of counting for the target words because headwords are usually the most frequent members of word families, and thus the most likely members to be known. This is supported by a corpus-based study which indicated that the headword was the most frequent member of 82% of the most frequent 1,000 word families in Nation's (2006) British National Corpus word lists (Brown, 2018). Using headwords also reflects the nature of L2 teaching and learning (Brown, 2018;Dang & Webb, 2016).…”

“…This is supported by a corpus-based study which indicated that the headword was the most frequent member of 82% of the most frequent 1,000 word families in Nation's (2006) British National Corpus word lists (Brown, 2018). Using headwords also reflects the nature of L2 teaching and learning (Brown, 2018;Dang & Webb, 2016). That is, L2 teachers and learners usually receive lists of headwords without their inflections and derivations, and, therefore, are most likely to choose headwords to teach and learn first.…”

Can Japanese learners of English comprehend inflectional and derivational forms in listening? Testing the validity of the word family counting unit

“…Fifth, this study did not assess knowledge of word forms having multiple affixes. Brown (2018) reported that among types in the first 5,000 WF6s, 26.9% (inclusive of base words) had two or more affixes. Considering that such word forms could be more difficult to understand than words with a single affix, the present study may have overestimated learners’ ability to handle morphologically complex words in listening.…”

“…In their analysis of 21 written texts, Laufer and Cobb (2020) found that derivational forms accounted for 7.78%, 7.88%, 5.04%, and 3.17% of the tokens in academic texts, newspaper articles, classic novels, and graded readers, respectively. Brown (2018) found that derivational forms comprised 13.4% of the coverage provided by the first five 1,000‐word bands of Nation's (2006a) BNC wordlists. We will use the figures from these two studies, together with data from McLean (2018) and Ward and Chuenjundaeng (2009) as previously presented, to estimate the amount of lost coverage due to unknown derivational forms in various discourse types.…”

“… Note . These estimates were calculated by multiplying the percentage of tokens that were derivational forms in (a) the first 5,000 words of the BNC according to Brown (2018) and (b) Laufer and Cobb's (2020) corpus by the percentage of unknown derivational forms in Ward & Chuenjundaeng (2009) and McLean (2018). …”