Interpreting service provision is good value for money

In this new episode of the Language on the Move Podcast, I spoke with Dr Jim Hlavac about interpreting in Australia.

Dr Hlavac is a senior lecturer in the Monash Intercultural Lab in the School of Languages, Literatures, Cultures and Linguistics. He is a NAATI-certified and practicing professional interpreter and translator. NAATI is Australia’s National Accreditation Authority for Translators and Interpreters.

Dr Hlavac’ research interests relate to interpreting in healthcare settings, interprofessional practice with trainee professionals with whom interpreters commonly work, and the incidence of interpreting and translation amongst multilinguals and in multilingual societies.

In the conversation we explore how professional interpreters, language mediators, and language brokers help to support fair and equitable access to healthcare and other forms of social participation.

How does interpreting work in practice in a hospital setting? Who gets to interpret? How is the need for an interpreter identified? Who pays? What is the role of policy vis-à-vis bottom-up practice? Is the process the same for all languages? Will AI make human interpreters superfluous?

Enjoy the show!

This is early days for the Language on the Move Podcast, so please support us by subscribing to our channel, leaving a 5-star review on your podcast app of choice, and recommending the Language on the Move Podcast and our partner the New Books Network to your students, colleagues, and friends.

Further reading

Beagley, J., Hlavac, J., & Zucchi, E. (2020). Patient length of stay, patient readmission rates and the provision of professional interpreting services in healthcare in Australia. Health & Social Care in the Community, 28(5), 1643-1650.
Hlavac, J. (2014). Participation roles of a language broker and the discourse of brokering: An analysis of English–Macedonian interactions. Journal of Pragmatics, 70, 52-67.
Hlavac, J. (2017). Brokers, dual-role mediators and professional interpreters: a discourse-based examination of mediated speech and the roles that linguistic mediators enact. The Translator, 23(2), 197-216.
Hlavac, J., Beagley, J., & Zucchi, E. (2018). Applications of policy and the advancement of patients’ health outcomes through interpreting services: data and viewpoints from a major public healthcare provider. The International Journal for Translation and Interpreting, 10(1), 111-136.
Hlavac, J., Gentile, A., Orlando, M., Zucchi, E., & Pappas, A. (2018). Translation as a sub-set of public and social policy and a consequence of multiculturalism: the provision of translation and interpreting services in Australia. International Journal of the Sociology of Language, 251, 55-88.
Long, K. M., Haines, T. P., Clifford, S., Sundram, S., Srikanth, V., Macindoe, R., Leung, W.-Y., Hlavac, J., & Enticott, J. (2022). English language proficiency and hospital admissions via the emergency department by aged care residents in Australia: A mixed-methods investigation. Health & Social Care in the Community, 30(6), e4006-e4019.

Transcript (created by Brynn Quick)

Dist Prof Piller: Welcome to the Language on the Move Podcast, a channel on the New Books Network. My name is Ingrid Piller, and I’m Distinguished Professor of Applied Linguistics at Macquarie University in Sydney.

My guest today is Dr Jim Hlavac. Dr Hlavac is a Senior Lecturer in Translation and Interpreting at Monash University in Melbourne. Today we’re going to talk about language barriers in a diverse society and how they can be bridged through interpreting between different languages. Welcome to the show, Jim.

Dr Hlavac: Thank you very much for the invitation, Ingrid, and to be on the Language on the Move Podcast.

Dist Prof Piller: Maybe I should say servus and tell our listeners – Jim and I are old friends, and usually we would have this conversation in German because that is our main shared language. So, doing this in English is actually a bit unusual for us. Maybe, Jim, you can tell us a little bit about yourself. How did you get into interpreting?

Dr Hlavac: Well, Ingrid, it’s probably not uncommon for people in my situation to have been brought up bilingually, or with even three languages, but also mobility – living in different countries – being born in Australia but then going to the birthplace from my parents when I was 7.5. And then, going back to other places where I have relatives and friends, spending time in Europe growing up, then coming back to Australia. So, often mobility has been affected which has accounted for my acquisition of languages and also my use of them.

When I travelled again from Europe to Australia in 1995, I had done kind of ad hoc unpaid translation and interpreting work for others, and I decided I really should formalise my credentials. So, I attempted a test and passed it, and since then I’ve been a what’s called a NAATI – NAATI for those who don’t know – it’s the National Accreditation Authority for Translators and Interpreters – I’m a NAATI translator and interpreter, and I work across 3 languages – English, Croatian and German.

Dist Prof Piller: Thanks a lot, Jim. Jim, maybe can you tell us what a professional interpreter actually does? I don’t think everyone knows. I mean, it sounds very glamorous. What do you do?

Dr Hlavac: I’m glad it sounds glamorous. Some parts are glamorous, some parts are less glamorous. So, what you do if you’re a professional interpreter is that you should have training, which I do have. You should have credentials such as I have from NAATI. Basically, when you work with other people you are working with 2 or more people who don’t have another language. When you work with them, you interpret everything that they say or sign – everything – so you don’t leave things out. You don’t add things. You don’t distort things. You’re impartial. You’re neutral. You’re not on anyone’s side, regardless of who’s paying for you. If you do have a particular relationship with a particular party, that should be declared to the other one.

You also observe confidentiality. Often, interpreters work in situations where people are talking about quite personal or intimate details, and it’s important for an interpreter to observe that confidentiality and to not pass on to anyone else information of events happening in interpreting assignments.

Interpreters work sometimes on site face-to-face with others. Sometimes it’s remote by video interpreting facilities or telephone. We all know about COVID. Everything went remote. So, there are different modes that you can use to communicate with people. But that, in a nutshell, is what a professional interpreter does.

Dist Prof Piller: So, you’ve been stressing “professional” interpreter now, and I’m wondering about – I mean any bilingual can interpret, right? People who don’t have the qualifications you have can also go and interpret, so can you maybe tell us what’s the difference between a professional interpreter and a language mediator or language broker?

Dr Hlavac: So, Ingrid, lots of bilinguals do interpret. If you speak to some bilinguals, they’ll say, “I can’t interpret, and I hate having to do it,” so it’s not a natural progression. It is something else, but you’re right in that many bilinguals do, as a matter of course, do it within their families or circles of friends or whatever.

So, what distinguishes a professional interpreter from a mediator or a broker is the following. I’ve talked about a professional interpreter. A mediator typically is someone who has a different role. They might be a youth worker, a settlement worker, a social worker, housing worker or perhaps a guide at a hospital, etc. where their primary role is to do something else, i.e. to help a person find employment or housing or what have you. And they might do so using another language other than, let’s say, English in Australia, which is the dominant language. Sometimes they’re just having conversations in that language. Sometimes they might be working with an English speaker as well, in which case they do interpret. But they often don’t know or care to know that when they work as an interpreter relaying other people’s speech or signing, that they have to do so fully without distortion. They can’t add their 2 cents’ worth, so to speak.

So, there’s always an issue with a mediator that their own primary role gets in the way, or they’re advancing the situation of a person for settlement or housing, and often the linguistic skills that they have are questionable. Sometimes they can be good, but they haven’t been tested. They don’t see themselves as an interpreter. They don’t know about ethics, etc.

A broker is something else. A broker is typically a family member who is often pressed into service. Sometimes they put their hand up, but often they’re pressed into service. Often, it’s a child who, if the parents don’t speak English, let’s take Australia, is there to interpret what the parents say to an English speaker and vice versa. Classic situations are hospitals, maybe police stations, other places, etc. Now, a broker is a family member, and so although they might look like a person doing interpreting, what they’re doing, their primary role, is being a family member. They’re looking after their parent, or whoever it is. They’re advocating for their interests. They’re making sure that what they hear and what they say is conveyed to their advantage. They’re also available all the time. They understand the parents’ language very well, etc. They’re also available all the time, and they’re free. So, they sound like they’re really great people to use in these situations, and often they are.

But there are some pitfalls, and the pitfalls are that not every child wants to or should be in that kind of a situation. A child can never typically tell a parent how to behave, what to do, because the power relations are such that they’re there to simply hear what they’re told to do.

There’s also many cases of brokers intentionally or unintentionally changing things. Imagine in a healthcare interaction the parent says something and the child doesn’t quite understand or really fully grasp what it’s about and says what they think the parents says. They convey that into English, and so what the healthcare worker hears is a description of symptoms that are actually different from what the parent says. Or, conversely, they might not understand the healthcare professional properly, be too shameful or kind of shy to ask for repetition or clarification, and they tell the parent something else that they think they’ve heard from the healthcare professional.

So that can lead to misdiagnosis, forms of treatment being misunderstood or not followed, to quite embarrassing situations. Let’s say an adult has a particular health issue which is an intimate issue. Is it appropriate that the child is privy to that information, and are they really likely to convey that? And also, when you think about yourself, would you like to go to the doctor and have your brother-in-law sitting next to you and you’re divulging information about your medical history and expecting your brother-in-law or whoever it is to recount this accurately and correctly, and they’re not going to change things that the doctor might say to them? How does that affect your relationship with your brother-in-law afterwards if he’s privy to all these things?

Dist Prof Piller: Yeah, look, I’m sure there are many, many difficult situations, and you’ve probably got a huge amount of stories to tell us. You’re not only an interpreter yourself, you’re also an interpreting researcher. A lot of the research you do is in healthcare, and you’ve already started us on healthcare. I guess, by the sound of it, it sounded like you’re not a huge fan of language brokering, and you pointed out all the problems that there are with family members actually interpreting for other family members.

But at the same time, we kind of know that it happens, and so I guess I’d be curious to hear from you specifically about interpreting and language mediation and language brokering in the healthcare system. What are the main barriers that patients in Australia who do not speak English, or who don’t speak English well, what kind of barriers do they face in accessing adequate healthcare in Australia?

Dr Hlavac: Typically, they have a number of barriers. There are often low levels of health literacy. They don’t know the health system in this country. They don’t know what services are available or that they’re entitled to. If they don’t speak English fluently, then they might not know that they’re entitled to an interpreter in most healthcare interactions that they’re likely to have. If they don’t know that, then they’re not going to ask, or ask a family member to ask on their behalf.

So, the challenge is for healthcare workers to recognise that a person is unable to communicate effectively in English and to offer or to organise an interpreter on their behalf. I’ve done some research, and even amongst those people who claim that they do know that health interpreting services are for free, it’s often the healthcare provider who still ends up providing them. And it sounds silly, or sounds obvious, but often people with so little English don’t know how to ask for an interpreter. They don’t even have those skills sometimes. And if you haven’t got effective communication, then, as you know, as the healthcare professional, they can’t work out what the symptoms are, what the level of health literacy is. They can’t work out a diagnosis and things like that.

Dist Prof Piller: So, who actually has to ask? I mean, you’re saying patients may not know they have the right to an interpreter, or they may not know how to ask. What’s the role of the healthcare professional, or how does – if I go to the doctor and I don’t speak any English, how does it actually work that an interpreter comes in? How is that decided, and what’s the process?

Dr Hlavac: So, the process is that, if you go into a large hospital, particularly in a metropolitan area like Sydney or Melbourne, you’re likely to have front of house staff who knows that this is one of the questions that they would ask as a regular feature when they’re addressing you for the first time through triage or whatever. Now, if you can functionally express yourself clearly, fluently, then they’re unlikely to ask you, but they still might. So, they’re obliged to ask this question, “Do you need an interpreter?” or “What language would you like your healthcare services provided to you in?”, which is a kind of optimal question, you know.

So, it’s up to them, and there’s a lot of cultural competence training happening in hospitals. There’s a lot of information that healthcare workers learn through professional development through their respective professional associations – how to work with interpreters. There’s a lot of skilling up that has happened across, particularly, hospitals. GP clinics are not so skilled up very much. I’m tracking data that’s looking at use of interpreters by GP clinics. It’s lower. Aged care facilities are also lower, so we do have variation. They key thing is, often it’s the front of house person to make the diagnosis. If they don’t, though, the healthcare professional can make the call that this person, this patient, needs an interpreter. So that’s how it usually happens.

The other challenge is, I mentioned health literacy and what have you. There’s a lot of information that’s been translated as well. I know we’re talking about interpreting mainly, Ingrid, but here in Victoria, I’m based in Melbourne, there’s the Victorian Health Translations website, which is 28,000 translations of material related to healthcare across 150 languages. There’s a lot of information out there to advise people about healthcare conditions, and one of the challenges is the discoverability of these resources. How do you get to them? They’re there, but how does the person for whom they are intended actually access them?

Dist Prof Piller: I’ve been wondering about that a lot, actually, because they’re usually organised by language, right? So, if you’re not good at spelling the Latin alphabet, or if you don’t know the name of your language in English, it’s really hard to find that information.

Dr Hlavac: It is. Typically, it’s a family member often, a younger family member, I did talk about brokers, who can lead them there. But they also need to know about this existing. So we do have a challenge in the accessibility of this information to people we want it to reach. When you do get to that site, you’ll find that there’s not just written text there. They’re moving now to audio files as a way of conveying information to people because we have a lot of data to tell us that this is the way people like to consume health information. Not through written text, but through an audio file. And there’s audio plus video. So, the repository of translations in Victoria does reflect people’s preferred ways of reading or gaining information in other languages. And it’s also quality checked.

There’s a lot of work happening recently of, firstly, the translations being checked and sampled amongst communities. And secondly, when healthcare departments or healthcare facilities are looking to compose a document in English, let’s say about Covid or whatever, that they actually involve translators at the stage where the plain English version is developed in the first place. It’s very helpful if you can have translators as part of the group, working on them, so that when the translations are then developed you don’t have the issues of “What does this mean? Let’s rephrase this”, etc. So, there is a lot of work happening in this area to optimise health translations. But we’ll go back to interpreting because I know that’s your focus.

Dist Prof Piller: Yeah, well look, I mean translation is fascinating too, and that leads me to another question. How do we actually know which languages are needed? We can go back to the clinics, so the receptionist establishes that this person needs an interpreter, but how do they find the right interpreter? Or, going back to your translations, how do we actually know in which languages do we need to make available information about a particular condition, for instance?

Dr Hlavac: The big hospitals collect data on not only interpreter requests, but the languages that are being requested, and they direct their resources to employing interpreters either in-house or freelance for those languages which are in demand. But they could have, you know, within the catchment area of northern health here in Melbourne, they service residents across 150 languages. They also have data from the ABS. Every 5 years we have the census.

So, we do have a fairly fine-grained idea in each municipality or local government are, what the profile is of the languages of the residents there, and also the level of English proficiency. The census data, the census has a question – “If English is not your language spoken at home, what is your level of proficiency in English?”, right? There are 2 gradings – “not at all” and “not well”. When residents tick those responses, that’s pretty indicative that those are people that will need an interpreter. So, we’ve got some demographic data. We’ve got data from hospitals themselves to know which languages are needed.

In terms of sourcing the interpreters, yes, Ingrid, this is a challenge because for bigger languages we do have an ok kind of cohort of interpreters to fall back on, but for new and emerging languages like Rohingya, when Rohingyas started to arrive say, 5, 6, 7 years ago, Chaldeans 15 or 20 years ago, we had to quickly develop testing for potential interpreters for those languages. Then getting them out to be able to work in communities. Often, it’s a kind of chicken and egg situation where you kind of approach people who are community leaders and ask them if they know of people who have good language skills who might have been doing this before migrating to Australia. And to locate people who have the attributes that you’re looking for in a potential interpreter and supporting them through training.

Dist Prof Piller: Yeah, I guess one problem that also kind of relates to named languages, you know. I mean, in my own research I’ve encountered people who’ve said they needed an Arabic interpreter, but they actually needed someone with Sudanese Arabic but then got someone with Lebanese Arabic and it was really, really difficult. The interpreter couldn’t really understand them. Or there have been all these media reports about the Yazidis in northern NSW who speak a variety of Kurdish but couldn’t really work with the Kurdish interpreters because their brand of Kurdish was quite different. So, I guess that’s an additional challenge.

Dr Hlavac: It is, and we do know about them. South Sudanese Arabic – there are 3 varieties of Kurdish that NAATI credentials. There are regular meetings, and I’ve been a part of them, between the language service providers who are at the coal face (Australian or British idiom for “front end” or “grassroots level”). They supply the interpreters, and they get together with NAATI, with the professional associations, and they say, “Hey, we’ve got this problem. We can’t find interpreters for this language. We’ve got a high incidence of people reporting this language, but they can’t understand the interpreters.” There are different varieties of Kurdish, etc. So, these things are fairly quickly made aware to the people who need to know about them, and we do respond accordingly.

Australia, through NAATI, is probably the only crediting organisation to have 3 varieties of Kurdish. And that’s simply because, as you said, there are Kurdish varieties that are mutually incomprehensible. And the whole thing of interpreting is that you need to be able to communicate effectively. If Lebanese Arabic interpreters aren’t able to communicate effectively with a South Sudanese Arabic speaker, the interpreter needs to inform the service provider, the English speaker, about this issue, that they are unable to communicate properly and that they need to rebook the assignment with a South Sudanese Arabic interpreter.

You do have speakers who might be speaking varieties that are not your primary one. You kind of, well you know about this very well, Ingrid, you practice accommodation. You try and work out how do they speak, you try and avoid things that are specific to your variety. I’m often working with Slovenians, who I don’t understand that well, and they, through misallocation that happens. If you really can’t understand that, the onus is on the interpreter to declare this issue, and for that assignment to be booked with the correct interpreter.

Dist Prof Piller: So, does that happen a lot? Like, you talk about misallocation. Is that a problem in the system, and then if I’m, I don’t know, I need to attend the emergency department, for instance. Maybe there is not a whole lot of time, actually, for people to find out what language I speak, and then to book and rebook, so how does that work?

Dr Hlavac: Yeah, it’s not easy, but there is infrastructure to address this. If you turn up to emergency and you’re incoherent or what have you, there are people at front of staff who will try to work out how much English you have, and if you don’t have English, what’s your language. They’ll often ask you anything – your country or language – in English, etc. It’s often possible for front of house staff to at least work out the language or the country of birth. Often, the country of birth does not coincide with the language, but that’s at least a piece of information that’s helpful for the front of house staff to start the process of locating an interpreter.

The free interpreting service is available 24/7. This is financed by the federal government. It’s free, so the healthcare facilities with emergency departments use this service, particularly after hours, and the ability to be able to locate and get an interpreter on the other end of the phone is not bad. The waiting time is usually between 3 and 5 minutes on average, which is not bad. There is a fair bit of infrastructure in place to address this issue.

People say, “This costs a lot of money”, etc. But if you look at the sums and if you look at the rates of misdiagnosis, healthcare workers not being able to communicate properly, the health effects, etc. and how much it costs the health system when these things happen – it’s much cheaper to pay for interpreting services that address the linguistic discordance in the first place.

Dist Prof Piller: Jim, you’ve got fantastic data, actually, on how the provision of interpreting services kind of reduces length of stay in hospital and how it reduces readmission rates for linguistically diverse people. So, really, this kind of value for money that our interpreting system gives Australian society – can you maybe talk us through that research and how interpreting really, you know, improves outcomes for people from non-English speaking backgrounds and overall lowers the burden on the Australian taxpayer if you will?

Dr Hlavac: So Ingrid, yeah, that was data that was collected by a colleague of mine, and friend, Emiliano Zucchi, based at Northern Health here in Melbourne. He tracked the use of interpreting services over 10 years. In those 10 years, interpreting services greatly expanded, as did the population in the area, but what we had happening was, and we can’t quite say it was only the interpreting services that resulted in lower length of stay in hospital and lower readmission rates. We’d need to do what’s called multivariate analysis to say that conclusively. But what we did see was that the increase in interpreting services co-occurred with these really good health outcomes – reducing the length of stay in hospital, lowering readmission rates – those are compelling reasons. They’re also reasons that hospital managers like to see. It’s not just the fact that patients and healthcare workers can communicate with each other optimally. There are great healthcare outcomes that have occurred or co-occurred with this happening.

Dist Prof Piller: Yeah, that’s really brilliant. I mean, we’ve already been talking about NAATI a lot and provisions in Australia. Our listeners come from all around the world, so I was wondering whether you could talk us through how Australia compares in terms of provisions for people who don’t speak English or don’t speak it well to other countries and the interpreting provisions and translation provisions available there?

Dr Hlavac: So, Australia compares favourably. I go back to really 1975 when they changed the macro policy, social policy of Australia, to introduce multiculturalism. If it wasn’t for multiculturalism, the flow on effect of that such as interpreting services would not be in this country to the extent that they are. So, Australia compares favourably in that throughout your provision of services acorss health, education, defence, employment, welfare – no matter what it is, each department has to have a multiculturalism policy, including linguistic diversity.

Part of linguistic diversity is the linguistic diversity of the government employees in that department, but also the people who use those services. So, when you’re unemployed and you need welfare assistance, the government department that you go to has to have a policy on providing interpreting services if you require them. Health is a big area, what I’ve mentioned. The courts, police, defence, tourism, etc. So, it’s actually built into the provision of all government services.

When you have money from government at federal and state level to support this, you can build up an infrastructure. When you don’t have the government support, it’s much harder. It’s much less prevalent and widespread, so that’s really the reason why Australia does compare favourably and why, compared to other countries, you do find, you know, a good service in terms of interpreting service and translation.

Dist Prof Piller: So, you’ve already spoken a lot about top-down and that the policies in Australia are really favourable, and the funding situation is quite favourable. Can you maybe talk us through bottom-up efforts? What needs to happen in institutions? Government can only do so much, you know. We need the policy framework in place, but at the same time at the institutional level, as you said earlier, people have to make things happen. There has to be a commitment to multilingualism and service provision for everyone and so on and so forth. I know that, from your research, you’ve also done a lot at the institutional level. Can you tell us a bit about what works and what doesn’t work?

Dr Hlavac: That example I gave before, when language service providers gather around a table to talk shop, to talk about what’s happening, what are our problems, issues, things we’re not doing well. That’s an example where people who are at the coal face do tell those people further up about what their gaps are and how they can be addressed. People aren’t short of suggestions. Now, sometimes those suggestions can’t always be addressed, but there’s this interchange of people at various levels that does characterise the system here which is pretty comprehensive.

If I go back to the 1970s though, when I was talking about multiculturalism being a key thing, there were people such as police officers complaining to their local members of parliament to say, “I can’t actually interview this potential witness because they don’t speak English and I don’t speak their language. They’re getting someone off the street to interpret. What are you going to do about it?”. You had doctors writing letters to say, “I can’t treat my patients. What are you going to do about it?”. When the country had actually gotten to a stage where they thought, “Ok, migration is an ongoing thing. This problem is not going to go away. How are we going to solve the problem?”. There were a lot of activists in that period coming up with lots of suggestions, and that’s how a lot of almost revolutionary things happened in that period. We’re fortunate we’ve had bipartisan support from both Liberal and Labor parties. Both sides of politics continue to support multiculturalism. So, interpreting services have not become a political football which can affect their future existence. So, that’s how things kind of panned out.

I’m sorry I’m not giving you a very good bottom-up example, but there’s a lot of interchange happening at many levels, and the system is kind of being fine-tuned, reviewed, and it’s open to lots of suggestions which are forthcoming from lots of people.

Dist Prof Piller: Yeah, look, I mean, that’s the democratic process, I guess, and it is encouraging to see it working. Now, I hear a lot of people currently coming forth with suggestions about AI and saying, you know, “We won’t need interpreting anymore in the very near future because AI is going to do it all for us,” and all those translation apps and so on and so forth. So, I have to ask that question, Jim. Are language technologies going to make human interpreting and translation superfluous?

Dr Hlavac: Ingrid, what a question! It might, one day, not tomorrow or the day after. With voice recognition technology which is the basis for technology understanding human talk and then being able to convert it into another language is really advancing, as we all know. We can turn on the captions function and that will probably give a pretty good rendition of what I’m saying and what you’re saying.

So, we’re speaking English, and hopefully we’re speaking standard English and speaking reasonably slowly and clearly, so voice recognition technology is good if you’re speaking a big language slowly, clearly, and a standard version of it. If you’re speaking a slow, standard version of another big language, you’re probably going to be able to use technology that is going to, I don’t know, probably interpret most of what is said correctly without too many mistakes and distortions. So, the technology is there, and it’s improving.

However, there’s two things. Most of the interpreting assignments that interpreters work in in this country is they’re working with people who typically don’t speak standard varieties who are often, particularly in health, they might be sick, distraught, unwell, unhappy, they don’t speak coherently. They don’t speak slowly. They don’t speak clearly enough. And so, the technology is not there to be able to pick up what they’re saying to then reliably be able to transfer it into English.

For the time being, the technology is not good enough to deal with the vast array of different varieties that people use in their vernaculars when they’re interacting with a healthcare worker. You need a lot of feeding of data from all sorts of languages, including colloquialisms, dialect, variation, etc. to have a voice recognition technology system that reliably can replace an interpreter. I don’t think it’s going to happen tomorrow or soon, but it might happen in 10 or 15 years, but it’s up to interpreters to work with this because there still needs for many things to be some sort of human overview, or at least supervision of this.

I’ve got a PhD student who’s testing voice recognition and using a tablet and asking interpreters, “Do you want to take notes like you normally do, or do you want to look at the tablet and see what the transcription looks like? When you interpret, is it easier from that or from your notes?”. So, there’s research happening.

The other thing is though, Ingrid, if the technology makes a mistake and there’s some sort of horrible outcome, who has liability for it? If you try and contact Google Translate and say, “Hey you made a mistake and this cost me $100 million. Can I sue you?”, you won’t get an answer, probably, because it is unclear who is responsible for that transfer of recorded speech from one language into another if you use automatic or neural translation technology. So, it’s a grey area, but we’re not going to be replaced tomorrow I don’t think.

Dist Prof Piller: Yeah, look, personally I don’t even think in 10-15 years. I mean, there is so much technology hype, and I guess I’m also interested in the dangers of that belief that at some point in the future interpreters will be replaced because, as you’ve pointed out, it’s the most vulnerable and the most high-stakes situations where technology actually fails. Technology is great if I need to get directions, if I’m a tourist somewhere and sort of in the leisurely, fun situation. Then it’s really, really good to have Google Translate or Google Lens or whatever. But if I’m in a vulnerable situation, a high-stake needs situation in healthcare, before the courts or whatnot, I think there is a real danger, actually, of thinking that this leisure and fun situation is somehow going to transfer to that situation where it really matters. Where we need human accountability. Where we need to make sure that it’s the right variety, it’s all those connotations that are there and so on and so forth as you’ve explained so beautifully.

Dr Hlavac: Yeah, things are developing. People might think, “Hey, I used it on holiday, why can’t I use it with my legal client here?”. There are some disclaimers and warnings out there. So, for example, Optus has a particular function where they can do speech recognition software, so you can speak, let’s say, German to someone. And at the other end of the telephone call, someone can speak Italian or Swahili or whatever. They said this is good for general communication only. They’ve kind of used the term “general communication”.

They do warn that this is not suitable for health or legal or high-risk situations. So, it’s often up to people to assess what the level of risk, particularly if there’s a miscommunication or mistranslation, what the consequences of that are. So, you know, the messages, as you said, it might be good in low-risk situation, but as soon as you have something at stake, you need to ask yourself questions. And human beings are a better evaluator of risks are. Human beings do make mistakes, but they are better in dealing with high-risk situations than what the technology has to offer us at the moment.

Like we say to our students, though, those interpreters who don’t work with interpreters will end up without a job, but those interpreters who do work with technology can look forward to continuing to have a job.

Dist Prof Piller: Well thanks a lot. I think that’s sort of a good note to end on actually. Thank you so much, Jim. And thanks for listening, everyone. If you enjoyed the show, please subscribe to our channel, leave a 5-star review on our podcast or on your podcast app of choice and recommend the Language on the Move podcast and our partner, the New Books Network, to your students, colleagues and friends.

Til next time!