Introducing Knowledge-based Vocabulary Lists (KVL)

Written by: Norbert Schmitt, Karen Dunn, Barry O’Sullivan, Laurence Anthony, Benjamin Kremmel
Published on: Nov 21, 2022

Photo credit: TungCheung/Adobe Stock

Vocabulary is probably the key type of knowledge necessary for any language use, because if words to express concepts are not known, all syntactic and discourse knowledge is of little use. This is true for both first languages (L1) and second languages (L2). First language vocabulary (at least in its spoken form) is mainly learned incidentally from the massive exposure which children receive, and much second language vocabulary can be learned this way as well (see the other articles in this TESOL Journal special issue). But in classrooms and materials, second language vocabulary is largely taught in an explicit manner. This means that second language pedagogy must have a way of prioritizing which words to teach from among the multitude available. (English is estimated to have about 10 million different individual words (Brysbaert, Stevens, Mandera, & Keuleers, 2016).

For the past hundred years or so, word lists have been used to inform this prioritization. These lists have largely been based on word frequency, that is, how often various words appear in (typically L1) discourse. These word lists have been used to identify the words that learners are likely to encounter, thus likely to need, and likely to know. However, frequency has its limitations when predicting which words might be known by L2 learners, particularly at the level of being able to produce those words (i.e., being able to recall and spell them accurately) (see Schmitt, Dunn, O’Sullivan, Anthony, & Kremmel, in press, for evidence of this). It is true that learners typically know more words in high-frequency bands than lower frequency bands; for example, learners will typically know more words in the first thousand frequency band (1K) than in the second thousand frequency band (2K). But while frequency ranking works relatively well for bands (at least up until about the 5K band), it does not work so well for individual words (e.g., we know that pencil and orange will likely be some of the early words learned by L2 learners, but appear as relatively lower frequency vocabulary on frequency lists).

This suggests the need for word lists based on direct tests of learner knowledge, rather than on frequency, which merely predicts imperfectly what learners might have learned from their language exposure. This Current Issues article introduces the Knowledge-based Vocabulary Lists (KVL) created to fill this need. The lists should be useful for pedagogical purposes in which it is beneficial to know whether learners are likely to be able to produce and correctly spell the words they know.

TESOL Journal