Tower of EU Babel

Stop Press!!! Google are to abandon and pull Google Translate by the end of the year.
Multi-lingualism is at once the bane and the aim of EU projects. How can you reach out to a collection of 27 countries with as many languages? The issue of translation is a key one in the UnderstandIT project, as it is in most EU projects, and in our forthcoming meeting we will be looking at whether and how technology can help us.

It seems relevant to be writing a piece about translation and communication across languages while sitting in Frankfurt airport waiting for a flight to Lisbon with examples all around me of people using a non-native language to communicate. I’m on my way to a meeting of the UnderstandIT project where we are testing out an approach to e-learning course development called Concurrent e-Design. The course being used as an exemplar is a development of the VITAE course, which seeks to train VET teachers in how to coach their colleagues to integrate ICT in their everyday practice. The idea is that after the VITAE course has been honed by the CCeD process, project partners will pilot the VITAE course in their own institutions in Italy, Lithuania, Norway & Portugal. And here’s the rub. The course has been developed and piloted in English. How will it translate into the other languages?

The course consists of tasks which can be translated but those tasks require the use of digital tools which are often in English, such as the blogging and podcasting tool Posterous or the screencasting tool Screenr. What are the options?

Don’t translate
There is an argument for not translating any of the course. If English is an international language, then most European professional people should be expected to operate in English. End of argument. However this doesn’t really hold water. Training should be attractive and motivational and if your language skills are poor you will never want to sign up for a course not delivered in your mother tongue. The problem here is mainly the productive skills; being expected to speak and write in English. Many VET teachers are former crafts and trades people who may not have had the opportunity to get to a high level of English. They have already made one major life transition from tradesperson to teacher and may lack the motivation to make another major leap into English as a lingua franca.
Only translate the tasks
This would mean running the course in the native language and translating all the tasks but not doing anything about the tools used in the course. This is a practical proposition in an international project such as ours because the expertise to translate lies in-house so to speak. It does mean however that we will have to be more or less agreed on which tools we want to present so that guides to these tools can be produced in the local language. These guides probably already exist in the case of the most well-known tools so then it would be just a matter of finding them and guiding course participants to these resources with relevant links. This Danish site for example explains many of the most popular Web 2.0 tools you would want to use in class.
Translate by hand
This means getting people to translate the course. The resources to do this for the tasks probably exists within the project network but if we look further into the future and start to think about when the course will need updating to use new tools (think for example of the demise of Dimdim & Ning as free services, Teachers TV which was closed down and delicious whose future hangs in the balance) then how cost effective will it be to keep translating? Professional translators are expensive. And one of the aims of the project is to promote discussion not only within each course group but also across groups in a Community of Practice. Will we employ a translator to translate these conversations across languages? Unlikely.
Translate by machine
An alternative is to use one of the many automatic translators which is available. Many people complain about the low quality of the translations but I can only say that I have been monitoring automatic translators for a number of years now and they have got better and better. They are invaluable for getting an idea of what a text is about even if you would not want to rely on the translation for something important about a point of law or medical diagnosis for example. They have, in short, become useful; for example my husband was able to use automatic translators to process correspondence between him and the German police regarding the recovery of stolen property. So what is on offer? The most widely known is Google Translate which you can use to translate a piece of text or a whole website. If you install the Google toolbar then you will be able to get translations even behind password protected pages. Another useful tool is Lingro which allows you to translate individual words on a page with a language with which you are otherwise quite familiar. Use of these tools does require you to be connected to the Internet while translating though. The video below shows how these functions work.

International conversation?
I think it is feasible to imagine a conversation across languages such as in our proposed VITAE Community of practice powered by the Google toolbar. What I like about the toolbar is that the translation happens almost seamlessly. In order for this to work you have to avoid the need for users to highlight and right click text before getting a translation which is where the toolbar scores. I must admit however that I have not tested Google Translate when presented with several different languages on the same page.
Non-text resources
As the Internet becomes more sophisticated we are increasingly presented with material in non-text format. Videos are becoming especially popular but there is no quick way to get these translated. Even here, however, there are developments. You can add subtitles to YouTube videos, though only if you own the video. The deservedly popular TED talks now come with crowd-sourced translations. You can add sub-titles to videos at dotsub where, for example, you will find the very useful CommonCraft videos subtitled in various languages. There is also the problem of text which is not text! This refers to graphics which include text as for example on the Screenr home page. Tools such as Google Translate cannot detect this text.
Local tools?
A final major consideration in localising the VITAE course is in the choice of digital tools we make. The most widely used tools tend to be those presented with an English interface and so the question becomes, do we present tools such as Voicethread which are leaders in their field or do we use a local alternative just because it is in the local language? Sometimes it will make sense to go with the local tool but often it will be those in English which have proved their value most. In those cases it is likely that there are already local language guides in their use. Another solution is to choose multi-lingual tools. Most Google tools are available in European languages such as Google Docs, Sites and YouTube. WordPress, is also multi-lingual. But what of niche products such as Screenr which are only available in English? Are we going to miss out on the advantages of video feedback and guidance or is it worth the effort of producing local language guides?

Language teachers hate the translation tools, claiming that they make their students lazy but I think they could make a major contribution in our project.

Postscript: During the meeting our Italian partners presented their solution for the Community of Practice which we have promised as part of the project. They have chosen Elgg and installed a translation tool which means that users from any EU country can have the content translated to their language. This is a very promising step forward to mutual understanding and pooling of experience and I’m looking forward to seeing how curious people are to cross language barriers in practice.