The conversation at the Round Table on African languages and 4IR suggested new ways of thinking

Should South African universities focus only on their own African languages? What is happening about the digitisation of indigenous language literature? Is interdisciplinarity essential for the success of the Fourth Industrial Revolution, and has 4IR been used to enable economic growth and transformation?

These questions were posed to the speakers at the recent Round Table on "African Languages in the Age of 4IR", and the answers helped synthesise much of the event's discussion, and suggested ways of thinking to take the ideas forward.

The Round Table was hosted by the Community of Practice for African Languages (CoPAL) of the Teaching and Learning Strategy Group of Universities South Africa (USAf), the umbrella body of the country's 26 public universities.

This is the Q&A, which has been edited for brevity and clarity.

QUESTION 1: South African universities should also teach at least one language common in Africa, such as Swahili. Should we focus on South African languages only?

ANSWERS:

Professor Mbulungeni Madiba

Professor Mbulungeni Madiba (left), Chairperson of CoPAL and Dean of Education at Stellenbosch University: I have always believed languages thrive in an ecosystem. So, the diversity on the continent requires us to not only focus on our own languages. We need to adopt a pan-African approach. Swahili should be one of the languages but I don't think any university in South Africa is teaching Swahili.

Dr Karen Calteaux (right), Research Group Leader of Digital Audio-Visual Technologies, which is part of the Next Generation Enterprises and Institutions (NGEI) Cluster at the Council for Scientific and Industrial Research (CSIR): I fully, fully agree. We have seen with the piloting of AwazeMed (a mobile app that provides translation into all South African languages to aid) communication between health workers and patients, one version for midwifery and obstetrics, and the other for COVID-19) that so many of the nurses and doctors are saying they can often find a way to translate between our local languages, even if it means quickly finding a nurse who speaks that language. But they say the big problem is so many people from other countries in Africa come to our facilities, and they are not managing to provide equitable health services to them because of language barriers.

Dr Karen Calteauxe

I see on the chat line that the University of KwaZulu-Natal (UKZN) is already teaches Swahili so perhaps I need to collaborate with them.

Professor Langa Khumalo, Facilitator of the Round Table, and Director of the South African Centre for Digital Language Resources (SADiLaR) at North-West University: Professor Rosemary Wildsmith-Cromarty, I know you're the one who has provided the detail that UKZN was teaching Kiswahili – two modules – and have a Kiswahili-isiZulu workbook, and there are mechanisms to bring those modules back.

QUESTION 2: How can we improve on the digitilisation of literature?

ANSWERS:

Dr Calteaux: I think we need to distinguish whether we are talking about digitisation or digitalisation. Digitisation means taking the text or hardcover and turning it into a digital, electronic format. Digitilisation is processing that information in that book in certain ways.

I know there's a big digitisation project at the University of Pretoria – they have a digitisation lab - but I think this happens at most universities' digitisation units.

If the question is about digitilising literature, there's already good progress in that direction. We have an early reader ebook where we put in descriptions of images so sight-disabled readers can also follow what is happening in the story and in the picture.

Professor Khumalo: Most of the classics, for example, isiZulu books, are no longer in print, so a process to digitise some would be very welcome.

Dr Vukosi Marivate

Dr Vukosi Marivate (left), Absa Chair of Data Science at the University of Pretoria: In machine learning, one challenge in terms of the data pipeline we use to train our artificial intelligence, is the availability of data that is representative, and also curated with care. Facebook recently acquired this large data set, which comes from Google's indexing of the internet, and they've got all this data in isiZulu and Setswana but some of the sources are very disturbing. Some of the early work that was written down...[he does not complete the sentence]..That's why machine learning researchers are trying to work with linguists. We have to understand why the creators did the translations, and even sometimes these digital versions. These translations end up inside the machine learning algorithms, but you don't want to interact with a chatbot and it starts giving you answers or interpretations that don't make any contemporary sense or are very biased.

These things should be part of the scholarship.

QUESTION 3: From the presentations, it is becoming evident that universities need to rethink the idea of locating knowledge generation within a discipline. Interdisciplinarity is important, and collaboration between information systems, engineering, linguistics and sociolinguistics, among others, is important for the success of 4IR. What are your views?

ANSWERS:

Dr Bhaso Ndendze

Dr Bhaso Ndendze (left), Research Director at the University of Johannesburg's Centre for Africa-China Studies: This is under constant discussion in a lot of university councils. I would agree emphatically. The link between the skills produced, and the skills required in industry, is one way we've been looking at curriculum in the Fourth Industrial Revolution.

Thinking across disciplines is necessary regarding the rollout and evaluation of the deployed technologies, in terms of the socioeconomic and cultural effects of these technologies. It's very much in keeping with the spirit of the times, and we would all be best served through groups and committees and projects that work across disciplines.

Dr Marivate: Some of these emerging technologies have already been there for a while, as Prof Daniel Mashao (Respondent of the Round Table, and Dean of the Faculty of Engineering at the University of Johannesburg) was talking about. Now they're just permeating across different disciplines.

Almost every university now talks about transdisciplinary or cross-disciplinary work, and we need to make sure we can enable researchers to do that. It's very hard to do transdisciplinary work because you end up being asked: "Why don't you fall into a nice little box, and be evaluated around that?". This is a challenge, especially for emerging researchers, because it's almost like you haven't earned those stripes yet.

But we can enable very welcoming communities within our spaces. When we started the Deep Learning Indaba (an organisation he chairs that aims to strengthen machine learning and AI in Africa through activities that include an annual gathering), we made sure we created a community that allowed people normally outside the field, such as philosophers and linguists, who want to interact in an honest way. In the same way I took time to understand philosophy.

If we allow that more inclusive approach you start asking questions such as : "How do I , as a machine learning person, make sure I bring a linguist onto my next project because I can see where they fit in?'', and linguists can do the same.

That's pretty much what Prof Mashao is saying: the world is going multidisciplinary; we just need to change the structures.

Professor Khumalo (right): And in the Masakhane Project (a natural language processing research endeavour for African languages that invites non-traditional researchers to help build translation models), we are breaking the frontiers within the higher education sector, working with people who are outside the academy to advance the development of African languages, especially while we're already working with industry, and other sectors. So, we have to reimagine the idea of where we draw knowledge.

Dr Langa Khumalo

Dr Marivate showed us an article written by over 10 people. In the humanities we quintessentially work as individuals in silos. We publish a paper all by ourselves. Co-publishing is regarded as something that diminishes our stature as academics.

We have to start thinking broadly and collaboratively because it enriches the knowledge that we are generating as scientists. This is exactly why 4IR is relevant for us; it is opening new frontiers of knowledge.

QUESTION 4: Can you indicate instances where 4IR has been used by illiterate communities to enable equity of access, economic growth and structural transformation?

ANSWERS:

Dr Marivate:
I think the work the CSIR has shown today, especially AwazeMed, is what you can actually do once you're able to build up a good momentum towards a goal. It's focused, a killer app.

We tend to look at it as this snowball of money that has been coming into machine learning. And if you use it to work on a local African language, the question starting to pop up after the honeymoon phase of being given the funding is: "When do we get to see these innovations actually changing society?". This is a question I think everybody on this webinar has to start asking. As we continue in this quest to situate more of these language innovations in terms of 4IR, we will be wondering if we have enough time to make an impact before we say it's easier, unfortunately, to just teach people to work in English.

We want to make available our translation systems on an online platform, even though they're still rough around the edges, so people can see where they fail. And they are failing. There's a reason Google Translate doesn't have Setswana and Sesotho; they are not yet at the level where they can actually put it out there without people being disappointed.

Professor Khumalo: The Department of Science and Innovation is demanding we create projects and innovations that are impactful to our communities. That's what this question is speaking to: is science, is innovation in technology really making us make a difference in the communities we come from?

This is the fifth and last piece in a series of five articles published from the Roundtable on African Languages in the Age of 4IR.

The organising body, the Community of Practice on the Teaching of African Languages (CoPAL), was formed in 2015 to enable academics and relevant other university staff members to collaborate, network and share knowledge on issues of common interest. CoPAL is just one of numerous communities of practice operating under USAf's banner.

In addition to influencing and contributing to national and institutional language policies, the CoPAL seeks to benchmark, develop, advocate for and share good practices and relevant information needed to advance the teaching of African Languages in schools and universities. It does so by contributing to teacher training initiatives for African Languages and to the development of approaches to the teaching of African Languages that use African Languages in the teaching process.

This CoP also seeks to enhance regional collaboration among African Languages scholars, as well as to actively contribute to the establishment of linguistic networks to ensure that information and common understandings are shared.

Written by Gillian Anstey, a freelance writer commissioned by Universities South Africa.


CoPAL Colloquium on Round table on African Languages in the Age of 4IR
Powered by NewsSite