Research Bites: The Relevance of the Academic Vocabulary List (AVL)

Durrant, P. (2016). To what extent is the Academic Vocabulary List relevant to university student writing?. English for Specific Purposes, 43, 49–61.

Durrant compares the Academic Vocabulary List (AVL, Gardner and Davies, 2014) to university writing in order to understand how academic vocabulary is actually represented in undergraduate and graduate writing.

The Wordlists

The AVL is a more updated version of the popular Academic Word List. There are some important differences between the two:

Academic Word List (Coxhead) Academic Vocabulary List (Gardner and Davies)
based on a 3.5-million word academic corpus based on 120-million word Corpus of Contemporary American
based on headwords without regard to different meanings caused by changes to word families based on lemmas (“headwords plus inflectionally-related forms”) to take into account the various meanings of world forms
based on General Service List of high frequency general English words which may contain words that also have academic uses (e.g. address) but are not included in the AWL not based on any pre-existing list

The Problem

Durrant’s research is to provide insight into just how relevant the AVL is. Some of the problems highlighted about wordlists are that vocabulary varies too much by discipline to have any list be of value. Another argument is that wordlists are more useful (insofar as they are actually useful) for reading texts, not producing them. In other words, their productive value is questionable.

The Research

The research compared the AVL word list to the British Academic Written English (BAWE) corpus. Durrant looked at overall use of the AVL, as well as variation by student level, discipline, and genre.

The Findings and Conclusion

  • The AVL accounts for about 34% of the lexical words in the BAWE
    • 20% of this is covered by only 313 words
    • The most frequent 32 AVL items account for 5% of the BAWE lexical items
  • The AVL accounts for slightly more usage as their academic levels rise
  • There is wide variation between disciplines
    • While the average for the entire AVL to account for 20% of the BAWE is 313 words, there is great variation by discipline
      • 106 words in architecture account for 20%
      • 1,312 words in classics account for 20%
      • The median is 194
    • There is some overlap between certain disciplines
      • For example, 40 words from the AVL account for 10% of words in linguistics and physics (17% of items are shared)
        • The three words that cover 5% of the BAWE in these disciplines are however, therefore, and theory
      • About 30% of AVL represents shared words which account for 20% of the BAWE
  • There is signficant but small variation between text genres

The Implications

A relatively small amount of AVL words represent a great deal of academic writing while about half have very little contribution in terms of coverage. However, the words that do contribute to a great deal of coverage vary by discipline. Durrant argues that these results may seem to imply discipline-specific vocabulary teaching is a warranted approach. Nevertheless, he argues that is usually not practical nor desirable “given the cross-disciplinary nature” of academia. Durrant recommends focusing on the most frequently overlapping words (427 lemmas) and then moving on to either more discipline-specific lists or, vocabulary strategies such as inferring meaning or skipping unknown words (here, he refers to Nation’s [2011] “Learning vocabulary in another language“).

I have adapted the word list from Durrant’s work into an Excel file. The file contains the most common academic words that are shared among 30 disciplines, sorted by part of speech and frequency. Please click here to download it.

2 thoughts on “Research Bites: The Relevance of the Academic Vocabulary List (AVL)

Leave a Reply