All Things EAP - Episode 3

[00:00:00] Laura Richards: Hi, and welcome to All Things EAP, a podcast produced by practitioners across the globe and hosted by BALEAP. It provides a fun and friendly look at every aspect of EAP, helping you keep up with the latest developments and building an inclusive and equitable community around the world. Following the BALEAP, we hope to be collaborative, professional, developmental, and inclusive.  Each episode, we'll dive into an aspect of EAP with discussions, reflections, and insights from practitioners in different contexts.  Whoever  and wherever you are,  we hope you find this podcast informative and insightful.

[00:00:55] Xiaoqing: Hello and welcome back to our EAP podcast. I'm Xiaoqing. 

[00:00:59] Sanchia: And I'm Sanchia.

[00:01:00] Xiaoqing: Today we continue our exploration of EAP by discussing machine translation and generative AI. 

[00:01:06] Sanchia: Ooh, a hot topic. 

[00:01:09] Xiaoqing: I'm happy to introduce my colleague Andy McIntosh, a language learner educator with over 25 years of experience. Andy's research focuses on the effects of translation app usage on text comprehension and vocabulary retention. He's currently a senior language lecturer at Xi'an Jiaotong Liverpool University. 

[00:01:28] Sanchia: We're also excited to welcome Mike Groves, an EAP educator and researcher who has been teaching EAP since 2007, and who began publishing on machine translation back in 2015. He's currently a senior lecturer at Lingnan University in Hong Kong. In this episode, Mike and Andy explore how students and staff are adapting to advancements in machine translation and generative AI. They also discuss some findings from Mike's recent research that might just challenge what we think we know about these tools. To what extent are they helpful aids? What are the implications for academic integrity? Listen on to find out.
. 
[00:02:13] Xiaoqing: Welcome to our podcast. How are you, Mike? How are you, Andy? 

[00:02:17] Mike: Yep, good.

[00:02:18] Andy: I've just been complaining about the weather. 

[00:02:22] Xiaoqing: Yeah, the weather is not ideal here in Suzhou. It has been raining for the last two weeks. What a perfect weather to talk about cheerful EAP. So in recent years, we all noticed that the use of machine translation tools like Google Translation has become very common among EAP students and also EAP practitioners. While these tools offer convenience and speed, there are varying opinions on their effectiveness. So in your teaching context, Mike and Andy, what are the attitudes of EAP practitioners and the students towards machine translation?

[00:03:02] Mike: Well, I think the students are using it quite a lot in Hong Kong and I think the EAP practitioners are not very engaged with it. So it would surprise me if within a class of 20 students, fewer than half were not using it on a regular basis. But it's not something that we've got in our regulations. It's not something we've got in our teaching materials. It's not very much discussed. It has changed a bit, since the advent of generative AI. and some policies are saying generative AI and machine translation. Others are just saying, AI, which is not very clear at all. so in summary, it hasn't been formalized. The policies and the attitudes are not really out in the open should I say.

[00:03:55] Andy: Would agree with what you say, Mike, the research I've been involved with here in Suzhou, has shown over the years that a huge majority of students use it in a variety of ways. I think they're still mostly focusing on reading writing, but as machine translation has grown more sophisticated, there is evidence that they are actually using it in the classroom to translate live speech, or basically the lecturer what they're saying from English to Chinese. They do have some concerns about the quality and also they know that basically the longer the text that they're trying to translate, the more likely they're going to have problems. But yeah, when we talk about the EAP lecturers and teachers, I think there's a lot of differences. Some are just, they don't know about it really. Others probably disapprove, but choose to ignore. And a select few, may actually try to encourage their students to use it in a critical way. But I think the point that Mike made about how we don't have any kind of formulated attitude, towards it is very true.

[00:05:06] Xiaoqing: Yeah, I can definitely relate to that. I have students who wrote their business proposition in Chinese and then used Baidu translation to produce them in English. Uh, for those who haven't heard of Baidu translation yet, it can be seen as a Chinese version of Google translation. Usually, in these types of essay, there are some awkward sentences that I need to ask a student to clarify. And then they show me their original essay in Chinese to explain it. To be honest, I always have, um, mixed feelings about this. But, just like Mike mentioned, it's not in my teaching regulations. So what I can see is that these students definitely need help with their English, with their language.

[00:05:53] Xiaoqing: Right. So we'll talk about the EAP teachers, their attitudes. Do you think the lecturers in other disciplines, I mean, they're not teaching languages or specifically EAP. Do you think they tend to have the same opinions or they hold different perspectives on the use of machine translation?

[00:06:10] Mike: Yeah, I mean in 2018, I think it was, just before Covid hit, my friend Klaus Mundt and I did some research in the UK. And we spoke to various academics at UK universities. Everybody from lecturer all the way up to professor. And the only selection criterion we had for them was they're not involved in language teaching. And almost all of them were very relaxed about the use of online translation. It wasn't entirely unanimous, but the general viewpoint was that as long as it's the student's own work, and as long as the students are demonstrating their knowledge and their analysis, then they were perfectly happy for them to use it. More so, I'd say, with reading. But also generally with writing. The only sort of caution that was expressed was the idea of students with a degree from these universities going out into the world reliant on translation and what that would do to the university brand internationally. But apart from that, the academic staff we spoke to would generally accepting.

[00:07:21] Xiaoqing: Yeah, you mentioned earlier that it's essential for students not to completely or to blindly rely on it. So how can we guide our students to use machine translation critically then? 

[00:07:35] Mike: I think guide is a very important word because we can only advise and offer advice to students. We can't force them into a particular way of doing things. But what I tell my students is that they should be using it as a facilitation for learning instead of a substitute for learning. And whenever they're using it, they should be thinking of the, question, what did I learn there? I've just used this translation software. Did I learn some new vocabulary? Or did they use it in order to correct the grammar in their writing? If so, how did they use it to move towards a stage where they don't need it? In the future to correct those mistakes. So it's all about the student asking themselves, am I using this constructively?

[00:08:28] Andy: Yeah. Yeah. I would entirely agree with that. As you said, Mike, we can encourage students to use it critically in the classroom. But what they do outside. Well, we don't know. And obviously, this does produce concern. I think it can be frightening to some teachers that students have more control. And when it's used wisely, it will enable the students to produce better work, and this seems a good thing. And we've got to show them that their input is essential. It's part of them developing their own academic voice. If everybody just translates from language one to language two using Google. Then I get a lot of somewhat bland outputs. But we should be encouraging students to actually to question what they're doing. Is it for vocabulary? Is it for phrases? Because there are tools and ways of doing this in the machine translation. Um, I think most universities would now tell students. Oh, yeah, Grammarly. That's fine. Go and do that. And it can cause problems when students think, well, if Grammarly is okay, surely I can just write it in Chinese and then translate it all and put it into Grammarly this raises serious questions about who actually produced these words, and can they explain it if they're questioned on it? So, at the moment, we just tend to let it go. But I think it students should be encouraged to realize its potential, uh, but also their own potential.

[00:09:59] Xiaoqing: Andy, you mentioned something interesting earlier about difficulty of seeing or controlling what tools students use outside of the classroom. I have a question as I'm very interested in your perspectives. For tools that students can use outside the classroom, such as generative AI tools like ChatGPT, how would you compare these tools with machine translation in terms of EAP teaching and learning? 

[00:10:28] Andy: This is the big question. Isn't it? Okay. So generative AI, chat GPT, has come along recently and, grabbed all the headlines really. But for me there's a big difference in that machine translation, a student can write, let's say a paragraph in Chinese. So they've done that first part of writing. Generative AI, it's called Generative because the students write a prompt saying, write an introduction for an essay, and then they can copy and paste the teacher's essay instructions and AI gives them that introduction. So for me, that prompt is not actually a student creating work. You can argue they're creating a programming aspect, but machine translation doesn't create, whereas generative AI does. 

[00:11:21] Mike: Yeah, absolutely. And, Google translate or Baidu translate, you have to have something to put into it to get texts out, whereas as chat GPT or copilot, you, you ask them a question. I think there's two similarities that are quite important to think about. Firstly, a lot of people think that they're going to put us out of business. because of chat GPT, why do we need to have universities anymore? Why do we need to write essays anymore? And people have said that about machine translation, but it hasn't put us out of business. What it's done is shifted the focus of what we do. We don't spend as much time now looking at dependent clauses and the difference between although and however, because we don't need to, and I think that's actually quite a positive thing. How this will play out with generative AI and what happens at universities I don't know. But I don't think it's unreasonable to expect something similar that will happen that it will cause a shift in the intellectual focus of what students do rather than a replacement of it. The second similarity is that generative AI and machine translation actually level the playing field.. If you're in an English speaking university and your first language isn't English, you are at a disadvantage. You have to work harder than everybody else. An online translation and chat GPT can help level the playing fields. And I think that's a, that's a good thing. . We don't get to the level playing fields without thinking really, really carefully about our teaching and our assessment and all those kinds of other things, but I think the potential is there for both of these technologies to make university a little bit fairer for people whose first language isn't the language of that university and also for issues such as dyslexia and neurodiversity.

[00:13:18] Andy: It can certainly, as you say, it can help students and it can enable them to focus more on the content rather than the language being a barrier, both to their understanding and to what they're able to produce. I think you've said before, Mike, that if you know that the technology can produce language that bit better than you can, then of course you're going to want to use it. It can save you months of trying to up yourself to an IELTS level six or something. But, so in the long term, we'd hope that technology will mean that everybody is, more equal. And I like the idea that, it can help people who have either physical or mental limitations can help them to contribute to a wider society. 

[00:14:12] Xiaoqing: yeah, definitely. That's a very interesting point. It just reminds me of my time doing my master's in the UK. Obviously, there was no chat GPT at that time, and some of my teachers recorded their lectures for us. In the beginning, when I was still trying to get used to the all English lessons, I was very slow. So I went back home and listened to those recordings as if I were having a listening test, checking all the specific terminologies that the lecturer used with translation software. Most of the time, I realized that I knew the concepts in Chinese, I just didn't know how they were expressed in English. So machine translation really helped me a lot at that time.

[00:14:58] Xiaoqing: Right, so we talked about how helpful machine translation can be, but it also raises several ethical issues, for example, plagiarism, which all teachers are very sensitive about. So what ethical considerations should EAP teachers and students be aware of when using machine translation, and how can this impact the teaching and learning process?

[00:15:23] Mike: Yeah, that's a, that's a really interesting question. And actually for me, this is something that institutions need to think about more than students. And I know that some people who work in the translation field are very unhappy about Google Translate because they see it as kind of working off the back of the hard work of generations and generations of translators. That's not to be dismissed, but I don't think that's necessarily a concern for a second year undergraduate student. I've seen some things that have made me slightly uncomfortable in the use of this. For example, teachers say you must not use online translation during this course at all. And I think that's, that's questionable for two reasons. Firstly, because you're not preparing them for the real world. Secondly, there's no way to police it. There's no way to know if a student has used online translation or not. And if you say to the students, you must not use it. The honest ones will go, okay, I'm not going to use it. The more, should we say creative people with creative attitudes to rules will use it. And then they will be advantaged by it. You so you're saying you're advantaging students who break the guidelines and I can't see that that is in any way fair the other thing that i've seen that makes me feel very uncomfortable is teachers kind of or even higher up in the hierarchy people's saying, I can tell the students to use this. I can tell, that there's been some kind of misdemeanor happening and they're doing it with very little evidence. They might say my student three weeks ago, we, we gave him a piece of writing in class and compare it to what he's just submitted through email. Look at the difference. And that is, to me, is very, very difficult because A, there's very different writing , in a classroom and writing at home where you've got all the time you want in the world. Secondly, student does some writing in class. You say to them, this writing needs improvement in a lot of areas. A month later, they come back to you with improved writing and you say, I think you're cheating. It's really, really ethically a difficult thing. So, my answer to that is A, be honest about online translation with students and B, have assessments where they can use it and have assessments where they can't use it. And if you don't want them to use it, make that part of the assessment design. So, yeah, I think we need to really think carefully about the way that the institutions deal with this as much as we do with the students. 

[00:18:08] Andy: You gave a a lot of interesting points there, Michael. But when it comes to the reading, obviously, ethically, We do generally accept that using translation for knowledge, for our inputs, that's been accepted in one form or another for quite a long time. But the output, that's what Mike was talking about, imagine if you're supervising , a master's dissertation and so a student comes along and they discuss ideas with you that they've got, they've developed and you say, okay, that sounds good. Why don't you write them up? And they can then write them up, but you know, because you've got a meeting and perhaps you've seen a basic outline. The language, I think we've got to be a bit more flexible with that unless, as Mike said, we are going the old fashioned way and it's okay. Right. Students, you have a pen and a paper and you got 50 minutes or and this is something that's The university here is moving towards the students work on a group project, and therefore, at the end, they produce a report, and then they individually get brought in for a spoken examination in which they have to explain what they did, how they develop the support, and will almost be cross examine. they have to defend and justify how they have produced this report and if they cannot then the suspicion is that, well, you haven't done this work okay. The, the writing, I don't think at this stage, it's still a plan that the writing won't be subjected to so much scrutiny, other than looking at the general layout and language. But as you say, Mike, there is a lot of institutes who make a lot of noise, but aren't particularly clear on what they consider acceptable or unacceptable. These can vary between departments, between different lecturers, and so it can send mixed signals to students. I would say that students in some way should also be asked, how do they feel? What are the ethics of using machine translation? So we should involve students in the process.

[00:20:28] Mike: I think also one other thing to add if I turn off Auto correct when I'm typing, it doesn't even look like English. My typing skills are so bad. My spelling is bad. Now, when I'm writing stuff, I've got auto correct. I've got spell check half the time. I don't even notice that AutoCorrect is changing things. So I'm using technological assistance to improve my writing. Why is it okay for me to use it? But a student whose first language is not English is not okay to use it. That's a question that I've never been able to find the answer to. And I think it's incredibly unfair. 

[00:21:13] Xiaoqing: Yeah, so it seems like not only the student, but also teachers need to be knowledgeable about machine translation, its functionalities, and its limitations. Do you think machine translation should be included in teachers professional development? If so, how? For example, how can we develop teachers literacy in machine translation?
[00:21:36] Mike: Yeah, so I'm going to say yes and no to that. I think Should AI and general technological assistance be included? Absolutely. I think if we just say online translation should be included in teacher training, then it's going to become obsolete quite quickly. I think looking five years down the line, I very much doubt that we're going to have discrete tools, one for translation, one for essay writing, one for summaries, like we do now. I think they're going to merge into something new. But certainly concepts of technological assistance, the concept of authorship in an AI age, the concept of facilitative and substitutive use of these technologies has to be absolutely central to teacher education. Um, certainly in EAP and probably in the wider academic community as well. 

[00:22:34] Andy: Yeah. I mean, I think it's now generally expected that if you work, uh, teaching at any level, then you will be able to use a computer and open a Word document and write a Word document and use email. And here in China, if you don't use WeChat, well, you're not going to have much of a social life. So this technology is just the norm. Okay. And if we consider that education is first of all, positive influence. on the way society develops, then teachers have got to have an understanding of how technology can be used for the good, and ideally they'll also have an understanding of how students might be tempted to use it, uh, which isn't necessarily best for the students, so they can help students to see this may be convenient, this may be quick and easy, but at the end of the day, you need to develop your brain, your brain is still not connected to that technology, although that may change before too long as well. So, yeah, I think generally, technology, we would accept it as part of the teacher development expectations and, as technology develops, so teachers need to develop with it. 

[00:23:52] Mike: There's also, if I can just add something on that just popped in my head. There's also the idea of trans languaging now trans languaging a lot of people think is a kind of a buzzword, but it's a very I think important concepts that as Language teachers now, it's not about monolingualism anymore, or at least it shouldn't be it's about Using all the available communicative resources that you have in your repertoire to make a communicative act. For bilingual students, it might be the different languages they speak, but it can also be the thing they're holding in their hand. They've got this digital resource. They've got this incredibly powerful digital resource and why not see that as a resource for developing communication competence rather than a enemy or a subversion or a bad thing. And I think that needs to go into teacher development as well, that this is a resource. This helps our students. It might not make us feel comfortable. But it helps our students. 

[00:25:02] Xiaoqing: , it is interesting that you mentioned about translanguaging. Maybe we can talk a little bit about non standard variants of English. I mean, the global use of English has changed, and it has definitely led to the emergence of diverse varieties of language. Personally, I experience this a lot. Since most of my students are Chinese and we share the same first language, so sometimes it is very funny that students use some Chinglish, but we understand each other immediately. So, do you think should EAP teachers be more accepting of non standard variants of global Englishes? And how can machine translation deal with this?

[00:25:45] Andy: I think to a certain extent we should. There's a difference here between what I would call, the formal and informal varieties of English. So sure, when you're chatting away, you might throw in some colloquial terms, join your friends or so on. However, if you are producing a written essay, or even giving a presentation, then there's an expectation that it will be more formal in many ways. However, if you know the majority of your audience, and they speak a very similar variety of English to you, then surely using that English will aid communication.

[00:26:31] Andy: There's an interesting question, well, does a global world accept global Englishes? And machine translation, I think you can tell it whether you want American or British or Australian English. But I would hope, and maybe AI can come in here as well, that it will allow students to, to adapt, to play around with both what they put in and what the output is so that it represents part of their identity as both a language learner, and a student producing work.

[00:27:10] Mike: For me, I mean, when we talk about standard, non standard Englishes, I have to start off by saying that I am a middle class white man from Southeast England. So, you know, the world is designed for people like me. I think that acrolect is the word, isn't it? I have the accent and the standard English. I think, actually. It takes an enormous amount of work for students to align themselves to these kind of unspoken norms, whether that's accent, whether that's , a particular grammatical way of speaking. From my time in Malaysia, a very simple thing is that in the Malay language, there isn't really a plural. So when people write in English, very often they don't use a plural and that is kind of a marker. Now when these people, if they want to go and study outside of Malaysia, that is seen as wrong English. And if they want to realign the English that they have been speaking since early childhood to the standards and the norms of a little rainy Island off the coast of Europe, it takes them an enormous amount of time and effort and heartbreak. People are judged on the standard of English that they speak. It's wrong, but it happens. I think the online translation is able to break that down a little bit. It comes back to being fair, to stopping the disadvantage that some of these students have. So in that sense, I'm all for it. I think it's a good thing.

[00:28:43] Xiaoqing: Very interesting. I'm going to ask a significant question as the last one for today. It is a rather broad one. What do you envision for the future of machine translation within the field of EAP?

[00:28:58] Mike: I think there are different futures that we can think about, that one way of looking at a very optimistic way of looking at it is to say we still do as much EAP. But we'll get to the cooler stuff quicker, so we won't spend so much time on noun phrases and dependent clauses, and we'll actually get to more nuanced ways of expressing critical caution or whatever it is. So we'll spend the same amount of time, we'll just go up the scale quicker. And then when students get into year one of university there'll be upper level than what they're doing now.

[00:29:40] Mike: Another less optimistic way of looking at it from an EAP perspective, but a more optimistic way of looking at it from a marketing perspective is that we say, okay, look, students rely on this. They use it. It's good. When they're in their degrees in their departments, they can use whatever technology they want. We'll offer them support in how essays are written. But actually we will treat them like British students at a UK university. You start a year one you finish it year three. None of this preparatory stuff. It's cheaper. It's quicker. It will be attractive to some students. The future is somewhere between those two poles. There is a third thing to think about. And it comes back to lectures. Now, I don't know if you know this story about where lectures come from. It was from the middle ages before the printing press. Now, obviously, we have the printing press, we have the internet, things have changed, but we're still doing lectures. Universities change very, very slowly. Don't ask me to bet on the speed of this change, , it might be quick, it might be slow. I don't think it's possible to make really clear predictions about this, but it will change.

[00:30:49] Andy: Okay, I certainly agree that it can be more interesting to challenge students to organize their thoughts into the accepted format. And I think there's still going to be the need for human interaction. So it does seem old fashioned when you've got, I don't know, two or three hundred students in a lecture theater, and they're listening up there to the white middle aged man speaking his traditional variety of English. But often at the end, a lot of students want to go and they want to ask this professor questions. And I've been quite impressed by the fact that they'll come up and they'll ask questions in English, not always perfect English, but that desire for communication is still there. And I think this is something intrinsic to humans that, sure, we can use technology and all this, but we actually like the opportunity to interact with each other. And I think students still value that. 
[00:31:47] Mike: Yeah, it is actually a danger if we regard English language ability or any language ability as a kind of a way into certain social or societal situations. So you know, in the UK to get into a prestigious university needs a higher IELTS than to get into a less prestigious university. So the language is a gateway into certain social situations.

[00:32:15] Mike: And if people are learning to become dependent on online translation, whether that's through subtitles on glasses or earpieces or whatever it's going to be in the future, their communication is going to be digitally mediated. And I suspect that social aspects, gatekeeping aspects will then be limited to a privileged sector of society who is able to communicate without that support. So that's an argument that actually online translation and generative AI might actually make social divisions worse rather than better. Um, Which goes completely against what I said about 10 minutes ago, but um, i'm sure nobody will remember that. Yeah, that is a that is a danger In some ways, it will be a level playing field, but in some ways, it will go back to sort of a small elite speaking a language that is inaccessible to others.

[00:33:13] Xiaoqing: I think it's common for new technologies when they are first introduced to be accessible, mainly to those with more power and resources. But that's just a normal part of how technology develops. And I think that the goal is to use this advancement to make them available to everyone over time. Hopefully this will help create a more equal society in the end. Thank you so much, Mike and Andy, for sharing with us your valuable insights about this topic. Today, we mainly discussed machine translation and generative AI in the context of the UK and China from a language teaching and learning aspect. However, there are many more aspects to explore and there are many newly published research papers in other contexts as well.

[00:34:00] Xiaoqing: Therefore, we are making this a series, and this episode is part one. So if you have ideas about generative AI and would like to share your research with us, we welcome you to get in touch. 

[00:34:14] Laura Richards: Thank you so much for listening to this episode of All Things EAP. We really hope you've enjoyed it and that you've found something interesting or useful in it. We really want to hear from you as well. So if you'd like to share a comment, suggestion or an insight into something that we've talked about on one of the episodes, please send us an email or voice note at podcast@baleap.org, but you can also get in touch through X, formerly known as Twitter and our YouTube channel, which are both @AllThingsEAP. Thank you again for listening, and we hope that you can join us for the next episode. 

[00:34:49] Laura Richards: Take care. Bye-bye.