Neural machine translation in EFL classrooms: learners’ vocabulary improvement, immediate vocabulary retention and delayed vocabulary retention

Abstract

Neural Machine Translation (NMT) has gained increasing popularity among EFL learners as a CALL tool to improve vocabulary, and many learners have reported its helpfulness for vocabulary learning. However, while there has been some evidence suggesting NMT’s facilitative role in improving learners’ writing on the lexical level, no study has examined whether vocabulary improvement made with the aid of NMT leads to any vocabulary retention when the tool is no longer in use. The present study employed a quasi-experimental design to examine the extent to which editing with NMT may lead to vocabulary improvement, immediate vocabulary retention and delayed vocabulary retention in the legal context and how the findings may vary by learners’ proficiency level. ANOVA results revealed that learners with higher proficiency achieved a similar level of immediate and delayed retention compared to their vocabulary improvement, while the benefits of NMT for learners with lower proficiency were significantly confined to vocabulary improvement and immediate vocabulary retention.

Keywords:

• Neural machine translation (NMT)
• English as a Foreign Language (EFL)
• Google Translate
• vocabulary learning
• vocabulary improvement
• vocabulary retention

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Notes

1 While Google neural machine translation (GNMT) is only one type of NMT, many relevant studies have been referring to GNMT when investigating the effects of NMT, most likely due to its wide availability and prevalence among L2 learners.

2 In her study, Tsai used ‘GNMT’ to refer to Google neural machine translation, one type of NMT.

3 In this study, ‘editing with NMT’ refers to editing self-translated L2 text by comparing own work with NMT output.

4 The test assessed learners’ receptive and productive knowledge of high-frequency words at the 2000 level, i.e., the most frequent 2000 words (Nation, 2013) and was available at https://www.lextutor.ca/tests/levels/productive/

5 The words in parenthesis are the L1 form of the target words given to the participants as required lexical items for their L2 writing. 

6 This includes an English-Chinese legal glossary designated for court interpreters in New York City (https://ww2.nycourts.gov/sites/default/files/document/files/2018-05/translatingJustice_Chinese.pdf) and Minnesota (https://www.mncourts.gov/mncourtsgov/media/assets/documents/0/public/Interpreter_Program/Chinese_Glossary_Simplified_Characters.pdf) 

7 For consistency, participants were asked to use Google Translate-GNMT in this study as an NMT tool.

8 In this study, NMT served as a good model and CALL tool (Nino, 2009), as all target words were verified to be accurate. However, for educational purposes, upon completion of this study, the participants were reminded of the judicious nature of the study and the possibility of inaccurate MT output.

9 This scoring system was adopted from that used by Hulstijn and Laufer (2001). Although an incorrect response and a correct response due to prior knowledge have intrinsic differences, both were scored 0 in the sense that improvement from using NMT would not be applicable. However, In line with what was reported in the verification step performed before the experiment, there were no instances where participants showed pre-knowledge of the target words.

10 Since all target words had been confirmed unknown to the learners before the use of NMT, it was presumed that any successful improvement in learners’ L2 revisions was primarily attributed to the use of NMT.