Online translators for L2 writing: a comparison of student and teacher perspectives

. This short paper reports on the preliminary findings of a study into the use of Online Translators (OTs) by university language students. Students of Chinese, French, German, Japanese, and Spanish and their teachers responded to comparative surveys on their respective use and evaluation of OTs for L2 writing in formal language learning contexts. Findings indicate that teachers have little awareness of the range of strategies students apply when using OTs as writing tools. Concerns of OT misuse for cheating or as a replacement for language learning seem largely unfounded. Students, however, perceive a lack of guidance for the appropriate use of OTs. Preliminary findings suggest that teachers need to review their assumptions about students’ OT practices and that both students and teachers would benefit from technical and pedagogical OT training.


Introduction
OTs such as Google Translate and DeepL are freely available and are widely used for various translation purposes. Yet when it comes to formal language learning, scepticism about their reliability and suitability remains. Prohibiting or discouraging the use of OTs, many language teachers consider their use as a form of academic dishonesty (Clifford, Merschel, & Munné, 2013;Correa, 2014;Niño, 2009). However, the frequent and extensive use of OTs among language learners (Briggs, 2018) suggests that incorporating or at least accepting the use of OT in the language curriculum is inevitable (Groves & Mundt, 2021). Recent studies have, in fact, shown benefits of OT use in second language (L2) writing, such as improvement in lexical and grammatical accuracy, vocabulary development, and reduction of writing anxiety (Tsai, 2020). Increased confidence through OTsupported L2 writing can also help language learners develop positive writing strategies (Lee, 2020). Further, OT training can prevent indiscriminate OT use and dependency (O'Neill, 2019) and help language learners to become more selfreliant in their writing (Tsai, 2020).
This study investigates OT practices as reported by university language students and their teachers, focusing on the following research question: • to what extent do learner and teacher perspectives on OTs differ in terms of their use and appropriateness in formal L2 writing?

Method
This study collected quantitative and qualitative data via online surveys. Student data were gathered over two weeks using a combination of 25 closed and openended questions. Survey questions were generated in reference to studies by Niño (2009), Clifford et al. (2013), Briggs (2018), and Tsai (2020). One hundred and fifty-nine university students taking Chinese, French, German, Japanese, and Spanish at beginner, intermediate, and advanced levels participated in this study. Twelve instructors (representing all languages and levels) responded to a 16-question survey. STATA was used for the statistical analysis of the quantitative data and NVivo for the thematic analysis of the qualitative data. This article provides a preliminary analysis of the data, which will be explored in more detail in a following paper.

OT use: teachers vs students
Three of 12 participating language teachers confirmed familiarity with OTs, using mainly Google Translate for professional and personal purposes. Another three described themselves as occasional users, whereas the other half claimed no experience with OTs. One occasional user admitted that her knowledge about Google Translate was outdated. Taking the survey might have triggered her interest as she suggested, "these translators must have improved a lot recently. I should check them out". Those who considered themselves experienced explained that they used OTs regularly to communicate with colleagues, family and friends, collaborate with international colleagues, or translate their own literary work.
Students, however, seemed well familiar with Google Translate and a range of other OTs, such as DeepL or Papago. About 90% of students were using OTs in their L2, both on their computer (86.5%) or on their smartphones (80%). Frequencies ranged from daily (23.81%), a few times a week (56.35%), to less than once a week (23.81%).
As depicted in Figure 1, students reported they used OTs principally for reading (translating words or text passages) and to assist L2 writing, a practice anticipated by teachers. Responses also revealed unexpected OT use for listening and pronunciation, indicating that students were taking advantage of OT multimodal affordances. Figure 1. Purpose of OT use: teachers versus students

OTs for L2 writing
Only a few teachers provided suggestions on how their students might use OTs for L2 writing. They assumed that students were using OTs to do the work for them, "to get the assignment done quickly and with few errors and presenting it as their own" or by "writing first in L1 and then translate it into L2 to be used as assignment". Such assumptions reflect the widely held belief that OTs are used for cheating and are detrimental to language learning (Correa, 2014). One teacher speculated that students use OTs "to check and correct entire clauses they have first written themselves".
The analysis of the qualitative data indicates that students are using OTs to support their writing in a variety of ways, at different stages in the writing process, mainly: • as dictionaries: aware of the limitations of OTs for longer text passages, most students (87%) stated that they used OTs primarily for looking up words, spell checking, and improving a text stylistically by looking up synonyms and idioms; • as grammar checkers: students identified a range of OT capabilities to deal with grammatical uncertainties while writing, such as syntax, verb conjugation, gender, and adjective agreement; and • for proofreading: as one of the teachers suggested, students enter their L2 text into OTs to either confirm accuracy or get alternative suggestions.
These reports suggest that these students do not use OTs to bypass but to support their writing. As one student put it, OTs help "with specific words or grammar points but should not be used for whole sentences or paragraphs otherwise, you are hindering your learning". Their handling of OT output similarly suggests that they abstain from simply copying and pasting translated text in their writing assignments. To verify the accuracy of the output, they use strategies such as text manipulation (e.g. "putting smaller phrases from a bigger phrase and seeing if they are translated the same"), consult others (peers, teachers, or native speakers), or draw on additional resources (Google search, textbook, or dictionary).

Implications for formal language learning
Findings indicate that the majority of language students (about 89%) are using OTs, that they are aware of their limitations, and that they have a range of strategies to use OT to support their writing. Teachers, on the other hand, show little awareness or support of these practices. Half of them had never approached the topic of OT in class, and none included guidelines for OT use in their course outlines, leaving students guessing about the appropriateness of using OT for their written work (Figure 2).

Figure 2. Approval of OT use for L2 writing: teachers versus students
While divided or undecided on allowing students to use OTs, 92% of teachers agreed on the benefit of OT training to ensure 'ethical' use of OT. The findings of this study thus reinforce Ducar and Schocket's (2018) recommendation that teachers need to review their assumptions about their students' OT practices and their own knowledge of OTs. It further suggests that both students and teachers would benefit from technical and pedagogical OT training.

Conclusions
As students start to seize the affordances of OT for language learning, it is crucial that language teachers adjust their teaching practices and guide students to use translation tools appropriately. The data presented in this article indicates that current student practices can help inform a pedagogical framework towards ethical and effective OT use for formal L2 writing. Further investigations into OT practices of learners at different levels of proficiency and of different languages will be explored in the full study.