The Open Translation MOOC: creating online communities to transcend linguistic barriers
Department of Languages, Faculty of Education and Language Studies,
The Open University
Department of Languages, Faculty of Education and Language Studies,
The Open University
Department of Languages, Faculty of Education and Language Studies,
The Open University
Institute of Educational Technology,
The Open University
Department of Languages, Faculty of Education and Language Studies,
The Open University
Abstract: One of the main barriers to the reuse of Open Educational Resources (OER) is language (OLnet, 2009). OER may be available but in a language that users cannot access, so a preliminary step to reuse is their translation or localization. One of the obvious solutions to the vast effort required to translate OER is to crowd-source the translation, as exemplified by Wikipedia (Wikipedia Translation) or TED (Ted Open Translation Project).
From October to December 2012 the Department of Languages at the Open University UK ran a MOOC on Open Translation Tools and Practices (OT12). Participants explored a range of online open translation tools (Amara, Transifex, Google Translator Toolkit) that enable and facilitate the crowdsourcing of translation, dubbing and subtitling. For this MOOC, participants collaborated in the translation and subtitling of OER mainly to and from English and Spanish, but also Portuguese, French, Greek, German and Catalan. Forum discussions, synchronous online sessions, recorded podcasts and the tasks themselves were designed to provide input, facilitate discussion and share views not only on the linguistic specificities of translating different languages, but also on issues such as quality assurance in open translation and the ethics and practicalities of openness in education and translation.
Data for this paper is drawn from online surveys covering participants' backgrounds and their prior experience as translators; their expectations and motivation for participating in the OT12 MOOC; and their evaluation of the outcomes of the MOOC.
We understand MOOCs as events (Cormier, 2010) or, following the principles of connectivism, catalysts for starting conversations within a network (Downes, 2011), and therefore feel that they might be a suitable way of engaging online communities of translators, language teachers and learners, and those interested in the crowdsourcing of translations for OER. This paper attempts to open up a debate on how the world of open education can harness crowdsourcing and existing open translation models to further the openness agenda.
Keywords: MOOC, OER, Open Translation, crowdsourcing, open tools, open practices
This paper describes a MOOC in Open Translation Tools and Practices offered in 2012 by the Department of Languages at the Open University, UK (OUUK) and considers the role that this type of open course can play in bringing together volunteers interested in contributing to the translation of educational open content. The paper highlights the need to translate and localize open educational content, presents some examples of Open Educational Resources (OER) translation initiatives, and makes the case for crowdsourcing and an open translation approach as a viable solution to this challenge. The Open Translation MOOC and the expectations and experiences of its participants are described, and the paper concludes with some reflections on the suitability of MOOCs as community forming tools around OER projects.
The language barrier
Language is one of the main barriers to the reuse of OER (OLnet, 2009), although this problem is often underestimated particularly by speakers of English, a so-called global language. There is a tacit, though false, assumption when publishing openly in English that the rest of the world is able to access content in English. However, we know that English is the main language for only around 375 million people, whilst another billion approximately use it, with different degrees of proficiency, as a second or foreign language (Graddol, 2000). And whilst speakers of English as a second or foreign language might be able to read text-based OER, audio-visual materials such as recorded lectures or webinars can be considerably harder to understand. That still leaves around 80% of the world's population unable to access educational content published in English.
Thus, if OER are available in a different language from that spoken by potential users, a preliminary step to reuse might have to be their translation or localization (i.e. adapting the content to a particular region). This section describes the ways in which different initiatives have sought to tackle the issue of translating and adapting OER for reuse, and illustrates the important role that technology and crowdsourcing can play in this endeavour.
A small-scale example of translation and localization is that of UnisulVirtual, the distance education campus of UNISUL, a Brazilian university that started collaborating with the OUUK in 2007 to increase the number of courses that were offered openly in Brazilian Portuguese (Lane, McAndrew and Santos, 2009). UnisulVirtual translated a number of business and management, IT and digital multimedia courses from the OUUK's OpenLearn platform from English into Brazilian Portuguese, and also translated some of their own courses into English, which were then made available openly through OpenLearn's LabSpace (see UnisulVirtual courses).
MIT has partnered with different organisations to translate its Opencourseware (OCW) materials into Spanish and Portuguese (with Universia), Simplified Chinese (with China Open Resources for Education, CORE - a consortium of Chinese universities), Traditional Chinese, Thai, Persian, Turkish and Korean. The Creative Commons license used allows translation into another language as long as the license's requirements (BY-NC-SA) are met, and MIT has made it a prerequisite that a disclaimer is added to the translation to guard against legal challenges in case of translation inaccuracies. In the case of Traditional Chinese, it was the enthusiasm of Lucifer Chu, Chinese translator of The Lord of the Rings, that led to the creation in Taiwan in 2004 of the Opensource Opencourseware Prototype System (OOPS), a volunteer organization which set out to translate the MIT OCW materials into Chinese. Through online contacts and media coverage the organization attracted volunteer translators and extended its work to resources from other institutions, which led to a two year grant from the Hewlett Foundation in 2006. By 2007 more than 2200 translators from over 22 countries had completed the translation of over 500 of MIT's courses (Lee, Lin and Bonk, 2007).
Khan Academy, the non-profit organisation providing free educational resources mainly for science, maths and technology, has its own volunteer translation portal (like Wikipedia or TED Talks). However, as their resources are openly licensed, their translation into other languages does not need to take place with KhanAcademy involvement or approval. For example, Portugal Telecom has made KhanAcademy videos available in Portuguese for Portugal and Portuguese-speaking Africa. The KA Lite initiative, an offline version of KhanAcademy content developed by the team at the Foundation for Learning Equality, manages the translation of KhanAcademy content through the open source translation management tool Pootle in its KA Lite Translations site. In March 2013, KhanAcademy organised a Czech Translation Marathon bringing together 60 volunteers at the Faculty of Information Technology, Czech Technical University in Prague, to translate Khan Academy videos into Czech.
A similar initiative has been taken by open.michigan at the University of Michigan, US, which has called on the languages community worldwide to translate a number of microbiology and disaster management video resources created by the institution in collaboration with institutions in Ghana and East Africa respectively. The call for help points out that 'the [disaster management] lectures were designed to be used across the East Africa region, but their current English-only captions and narration make them largely inaccessible to regional French- and Swahili- speaking countries' (Ludewig Omollo, 2013). The campaign makes use of open tools such as Google Translate, YouTube and Amara to speed up the process, and has engaged local participants through a marathon translation competition or Translate-A-Bowl.
Wolfenden, Buckler and Keraro (2012) point out though that 'making content (OER) available to others is relatively easy; ensuring that it is useful is much more difficult but cultural appropriateness and task authenticity are essential to support learning' (p. 15). The following are two examples in which adaptation to local context is viewed as crucial to the reusability of content.
The TESSA project (Teacher Education in Sub Saharan Africa), initiated in 2005 and led by the OUUK, aimed to provide OER to support school-based teacher development and improve the quality of teaching in Sub Saharan Africa, where many teachers are unqualified or under qualified (TESSA, 2012). Its approach was to make available a bank of modular, flexible resources, originally written in English (as the working language of the project) but then translated to Arabic, French and Kiswahili and adapted to nine country contexts. Adapting OER to the cultural realities of the various partner countries and regions was seen as essential to ensure that the learning materials could support the authentic, situated experiences of the teachers.
TESSA materials were created by teams of authors from several countries in the project, and were structured to facilitate adaptation without having to rework all the material. A well-defined template was used and some generic sections were included, with others such as case studies being identified as more suitable to be adapted to each particular context. Wolfenden et al. (2012) describe the process of translation and adaptation, which was undertaken by teams of lecturers (subject specialists) recruited for that purpose, and propose the following typology of changes: a) changes to names of people and places; b) cultural and environmental changes (to fit local history and culturally relevant examples); and c) changes required to align the resource to the local curriculum (by adding or taking away material, or changing it to suit the level of pupils).
The process of adapting the 75 study units in TESSA took almost a year and a substantial amount of resource. However, the resulting changes are described as modest (in number and depth), possibly because "for many authors, making changes felt disrespectful to the original author, implying a criticism of the author's work" (ibid, p. 10).
Following the success of TESSA, the Department of International Development in the UK is currently funding the TESS-India project (Teacher Education through School-based Support in India), also led by the OUUK. Once again, the aim is to tackle the low achievement and high dropout rates in primary and secondary education in India by providing a network of freely available distance learning teacher education resources.
The resources are being developed by UK and Indian subject experts, and will be translated and adapted to the seven states in which they will be implemented. Three of the states share Hindi as their common language, although different dialects are spoken in each state, whilst in the other four states the working languages are Assamese, Bengali, Kannada and Oriya. There are many reasons, including political ones, for translating and adapting the content, not least the fact that teachers' proficiency in English is generally low, in spite of it being recognized as an additional official language of the Republic of India alongside Hindi.
TESS India is looking at different models to enable the translation and adaptation of the resources. Professional translations carried out by agencies have been rejected as too expensive and potentially resulting in content that would be written in a high register language unfamiliar to many primary and secondary school teachers in rural areas, who themselves often have very low qualifications. The TESSA model of employing linguistically proficient members of local existing academic subject communities to translate the resources from English is likely to be used for the translation to the official languages of the seven states involved in the project. The issue of Quality Assurance might be tackled through a central body, possibly a Non-Governmental Organisation, and carried out by ordinary users, rather than translation or subject specialist, who would check translations for linguistic accuracy and appropriateness. The possibility of crowdsourcing the translation of the resources into other local languages beyond the official ones in the project states is being investigated, and some links are being established with Wikimedia India to explore using their existing volunteer translation community.
Open translation describes the practice or discipline that develops at the intersection between open content, open source software and open production models (Hyde, 2009). It makes use of free/open software and open collaboration to engage a distributed volunteer workforce in the translation of resources that have been published openly on the web. The aim of open translation is to make resources available to the widest possible audience using open source tools, thus avoiding the creation of 'a critical bottleneck in the open knowledge ecosystem' (Hyde, 2009) that the use of proprietary software might impose. Its open production model encourages peer participation and draws on collective expertise and thus crowdsourcing, similar to the way in which Wikipedia has led to a re-evaluation of the role of the expert or Global Voices, a citizen media project, has turned citizens into journalists.
Crowdsourcing translation is already an established and successful solution to making content more accessible in some large-scale, high profile open projects such as Wikipedia (Wikipedia Translation) or TED talks (Ted Open Translation Project). In the project that forms the backdrop to this contribution we set out to assess whether:
In the following section we describe the Open Translation MOOC (OT12) and present some data on participants, their expectations and their evaluation of the experience.
The OT12 MOOC
OT12 was conceived as a MOOC, or massive open online course, and was developed and presented by the Department of Languages at the OUUK in the autumn of 2012. It lasted eight weeks and was organized as a traditional online course, with stated aims, weekly translation tasks and readings, webinars led by experts, and online discussion forums where participants could exchange ideas, support each other and seek help from the facilitators. Although there was the option of obtaining a certificate of participation (linked to the participant's online contributions), none of the activities were assessed formally.
The course tried to stay true to the openness in its title by using an open platform, LabSpace, part of the OU's OpenLearn site, and by publishing the course itself under a Creative Commons BY-NC-SA license. The resources used as background reading and for the translation tasks were OER, an approach that was guided both by practical considerations (copyright) and by the fact that one of the aims of the course was to introduce participants to open resources and open practices. The main aim of the course was to introduce participants to open translation practices and to some of the open tools that can be used to facilitate the open translation process, such as Google translation toolkit, translation workflow tools such as Transifex, and the video subtitling platform Amara. A further aim of the course was to explore whether a MOOC format would be appropriate to bring together a distributed community of potential or existing volunteer translators, and to point them to existing open translation communities they might want to join.
The OT12 MOOC included a range of activities and tools designed to promote discussion about open translation tools and practices and to enable participants to try to evaluate different translation tools by using them to do real translation tasks. We will describe four types of activities to give a flavour of what was included in the course: forum discussions, subtitling tasks, translation tasks and online seminars with guest experts.
a) Forum discussions
Throughout OT12 there was a course forum in which students were invited to discuss specific issues. The OT12 MOOC regularly referred participants to a FLOSS (free/libre open source software) manual on Open Translation Tools (Hyde, 2009). Several forum discussions were articulated around this reference text, and invited participants to comment, based on their experience.
An example of a forum discussion task from week 1 is given in Figure 1 below.
Figure 1: Forum discussion task on translation
Read the section Why Translate? in the OTT/FLOSS manual.
The authors of the OT12/FLOSS manual ask the question 'Why translate?' and provide a number of reasons.
Which do you think are the most important and why. Why do you translate?
If you use social media (as a blogger, or as a facebook/twitter user), do you translate any of your content? Why/why not?
Figure 2 shows a forum discussion task from week 2 in which participants are invited to share their expertise.
Figure 2: Forum discussion task on dictionaries
What online dictionaries or terminology databases do you use? The OTT FLOSS manual provides a very brief introduction to this topic, and includes some links to tools [...]
In this forum, we propose that you share an online (and preferably open!) dictionary or other terminology tool that you are familiar with, explaining why you think it's useful.
Please put the name of the tool as the subject of your message, and then provide a brief description of the tool, a link to it and, if you can, a brief critical review, perhaps highlighting some advantages and disadvantages.
If the tool you want to discuss has already been shared by someone else, just add your comments to that discussion.
b) Subtitling tasks
Part of the aim of the OT12 MOOC was to provide participants with the opportunity to try out open tools and engage in the joint translation of open resources. Two weeks were devoted to captioning and subtitling videos. Activities involved a discussion about the link between captioning and subtitling and accessibility, the participants' evaluations of automatic captioning vs. human captioning, and hands-on experience of using Amara, an open source platform for subtitling video content from the web. Although participants could caption or subtitle any video they wanted, we also created a group for the OT12 MOOC and uploaded a number of videos from the OUUK's OpenLearn site, so that participants could all work on the same ones and discuss the process as they went, in a dedicated forum.
c) Translation tasks
In terms of open translation tools, the aim of OT12 was to provide participants with the opportunity to work together on a translation using a translation workflow management tool. Participants translated a number of OER into their respective languages. The tool we selected was Transifex, which is widely used for the management of crowd- sourced translation and localisation of open projects. Transifex essentially breaks down a text into small strings to facilitate the translation, and members of the translation team can translate particular segments, and suggest corrections for those of others.
d) Online seminars with experts
During the OT12 course there were a number of synchronous, online seminars with experts. These were experienced translators and other professionals who worked with open translations tools or practices. Their interventions were scheduled to coincide with the work participants were engaged in that week. For instance, in week 2, when the tasks involved subtitling videos using the open platform Amara, the guest speaker was Jules Rincón, from Amara Community Support and Advocacy. The synchronous online seminar was recorded for those who could not attend in person, and the guest speaker was also available in an online forum for a few days before and after the seminar, to address any questions from participants.
Using Lane's tripartite classification of MOOCs (Lane, 2012), OT12 can be considered to be primarily a task-based MOOC (rather than a network-based/cMOOC, or a content-based/xMOOC), with an emphasis on skills acquisition based on the completion of a series of tasks. The network aspect was also salient, but for most participants interaction took second place to trying out the tools and engaging in task execution. The facilitators were upfront about the exploratory character of the course, and the 'learning by doing' approach that underpinned OT12.
Participants: expectations and experiences
There were around 300 active students in the first week of the course, although nearly 600 had registered and received biweekly digests and updates from the facilitators during the whole course.
Participants were asked to complete several questionnaires during the course, placed in the study calendar in the relevant weeks. Given the open, exploratory nature of the course, it is impossible to establish the response rate for each of the questionnaires, as the number of participants actively engaging with activities varied from week to week.
A language profile questionnaire (n=196) completed during the first week of the course provided the following data: the majority of participants had Spanish (32%) or English (22%) as their main language, with other sizeable linguistic minorities present (Brazilian Portuguese, 11%; Greek, 9%; French, 7% and Italian, 6%). Most participants were expecting to translate into either English (59%) or Spanish (24%), and almost 70% of them claimed to be highly proficient in their second language. In terms of their familiarity with translation, almost all respondents claimed to have experience of translation, either professionally (43%) or informally (48%), with only 8% of respondents having no previous experience of translation.
Participants were asked to fill in a questionnaire during week one stating their expectations for the course and providing some information about their motivation and previous experience of online learning. According to the survey data (n=56) the majority of respondents (73%) joined the course to learn more about translation. Only a minority had a specific interest in open tools and resources or in second language learning. Most had never taken part in a MOOC and in fact over half of respondents had not come across the term before they engaged with OT12. Only 3 respondents had previous experience of online learning. Nearly two thirds of respondents saw managing their time or their workload as the main obstacle to participation in the course. For a fifth of respondents the main challenge envisaged was the technology and for six of them, it was working collaboratively, either because of a preference for working individually or because of anxiety caused by lack of knowledge of the topic. Asked to identify what would be a successful outcome of their participation in the MOOC, most respondents mentioned gains in knowledge related to translation; although some were more specific and hoped that completing OT12 would enhance their professional profile. For a few the main benefit was in the connections and networks that could be forged through participation in the course. A large percentage of respondents were vague in their answers, mentioning 'completion of the course' and 'learning' as satisfactory indicators of success.
During the last week of the course participants were directed to a short evaluation questionnaire (n=35). Most respondents felt their expectations for the course had been fully (46%) or partly (51%) met. The additional comments were divided between those who felt they had had insufficient time to dedicate to the course and those who were not entirely satisfied with some aspect of the course such as level of guidance, lack of support from peers, or more emphasis than expected on the 'openness' aspect. Respondents had enjoyed learning about open tools and resources, taking part in the webinars and sharing and networking with others. They had been less happy with their own lack of time to work on the course and, in some cases, with particular aspects of the course like length of tasks or lack of personalized feedback. In spite of the initial anxieties, only one respondent mentioned problems with the technology. And paradoxically when asked to suggest improvements to the course, many respondents wanted more content and tasks, in spite of recognizing time pressures as one of the biggest challenges.
Although the overwhelming majority of respondents (85%) considered the need to collaborate online as a positive aspect of the course, many acknowledged that participants could have been more proactive in this respect. Yet, for many this was their first experience of online learning. The main comment in relation to collaboration was that sharing ideas and benefitting from the knowledge of others was positive.
When asked what they had gained most from the course, 82% and 76% respectively selected 'awareness of open educational resources' and 'better understanding of translation tools'. The chance to translate and network with like-minded people was also appreciated by about half of respondents.
OT12 and Crowdsourcing
As mentioned above, crowdsourcing translation is a well-established solution to making content accessible in other languages. In OT12, we wanted to evaluate whether an online community of volunteer translators could be assembled to translate open content within the format of a MOOC. In this section we reflect on the extent to which OT12 was a successful crowdsourcing endeavour.
Rather than a single strategy, crowdsourcing is an "umbrella term for a highly varied group of approaches" (Howe, 2009). Unsurprisingly, their lowest common denominator is the fact that they all depend on some form of contribution from the crowd. These contributions, however, tend to vary considerably depending on what is the ultimate aim of the crowdsourcing endeavour. Potential activities range from calling upon customers for the design of a new product to setting up a platform for bloggers. Accordingly Howe (2009) distinguishes four main types of crowdsourcing: crowd wisdom, crowd creation, crowd voting and crowd funding. OT12 falls into the category of crowd creation, which relies on the crowd's creative energy. Howe lists ten basic principles of crowdsourcing. We can claim to have followed most of these in OT12. Choosing the right model (Principle 1) is the first of them.
Once the aim of OT12 had been established, i.e. exploring the potential of open translation tools and practices to aid the crowdsourcing of OER, the right crowd needed to be addressed (Principle 2: Pick the right crowd). A holding page was set up and an initial message inviting people to participate was crafted and distributed to Computer Assisted Language Learning (CALL), Computer Mediated Communication (CMC), Translation and OER listservs the project team members belong to. Yet, as Howe (2009) correctly reminds us, while it is relatively easy to attract a crowd, it is considerably more challenging to keep them (Principle 3: Offer the right incentives). Thus it is important to have an understanding of the motivation behind people's contributions : "Personal glory, the chance to interact with like-minded peers, and the opportunity to improve their skills or simply to learn something new all play a role" (Howe, 2009, p.282). To a certain degree, this echoes Downes' (2012) conceptualization of MOOCs as a sandpit for highly self-motivated learners who need to be put in touch with like-minded peers:
"What we are trying to do with a MOOC is to create an environment where people who are more advanced reasoners, thinkers, motivators, arguers, and educators can practice their skills in a public way by interacting with each other. In such an environment, people can learn by watching and joining in." (Downes 2012)
All of the aforementioned factors were evident in OT12. 73% of the participants indicated in the pre-course expectations survey (n=56) that their primary motivation was to learn more about translation, whether this meant learning to use translation tools or simply to develop translation skills. Figure 3 gives an overview of participants' motivations for joining OT12.
Figure 3: Learning aspirations of OT12 participants
When asked what they perceived as a successful outcome of their participation in the MOOC, a minority (13%) saw the main benefit of joining OT12 as networking with other translators. A similar proportion was motivated less by self-interest than by reasons of sociality. They expressed the wish above all to belong and contribute to a learning community.
17% (n=9) made it clear that they wanted participation in the OT12 MOOC to enhance their profile in the world of professional translation ("personal glory"), either by leading to the conferral of a diploma, or by furnishing a positive evaluation of their work.
The minority for whom sociality seemed to have been the main incentive points us to an important aspect of Principle 3 (Offer the right incentives), which is closely related to Principle 10 (Ask not what the crowd can do for you, but what you can do for the crowd). It is reflected in Howe's (2009, p.288) observation that "successful crowdsourcing involves satisfying the uppermost tier on Maslow's hierarchy of needs", which leads him to conclude that "people are drawn to participate because some psychological, social, or emotional need is being met".
OT12 was set up using Howe's "benevolent dictator principle" (Principle 5: The dumbness of crowds, or the benevolent dictator), which states that people's contributions need to be directed. "The most successful crowdsourcing efforts", he explains, "are products of a robust collaboration between the crowd and the individuals guiding them". The "load" of guiding was distributed among the 5 project team members with one person working as the main facilitator in the forums during the entire project period. There was a "Things to do this week" introduction, a midweek summary of contributions, and a "Summing up week X" digest for each of the eight project weeks. This approach allowed participants to dip in and out of OT12 and to engage only with those activities they were interested in, and/or to simply keep in touch with OT12 developments even when they were unable to make a visible contribution in a given week. It allowed the project team to cater for fluctuations in participant motivation and other demands on the crowd's time.
At the same time the team experienced "the community is always right" rule (Principle 9): the language pair originally envisaged for the translation work was English and Spanish, and OT12 had been advertised as such. Yet many participants worked within other language pairs, more specifically Brazilian Portuguese/English, and French/English. In order to respond to this unexpected reality, space for different translation subgroups was provided. "You can try and guide the community", as Howe (2009, p. 287) puts it, "but ultimately you'll wind up following them". While this could be interpreted as contradicting the need for guidance which underpins Principle 5, it also explains the use of the term "benevolent", i.e. someone or - in the case of OT12 - a team of people who provide a course structure or, in Howe's (2009) words, "a layer of administration", and who are there to answer the crowd's questions. An explicit goal of OT12 was to be exploratory though. As a result, the facilitators relied heavily on the knowledge and expertise of the participants.
The fact that people are busy leads to Howe's "keep it simple and break it down" rule (Principle 6) which was adhered to in OT12's modular approach based on weekly tasks and the use of bespoke fora linked to the execution of these tasks. Again, this set up also enabled participants to contribute only to those activities they were genuinely interested in (see above), and illustrates Howe's exhortation in agreement with Benkler (2006, p. 285): "While creative capacity and judgement are universally distributed in a population, available time and attention are not". It is therefore equally important "to keep the nature of the task simple. […] By bringing clarity and simplicity to your appeal - find the best minor league pitcher, translate this paragraph into French - you greatly increase the odds that someone will want to contribute " (Howe 2009, p. 286).
On a much smaller level, Principle 6 is also reflected in the use of the translation workflow tool Transifex, a platform for managing the crowdsourcing of text-based translations. It is a flexible version-control system for translation strings, which means that files are broken up into smaller segments to facilitate their translation. "Any task worth doing", as Howe (2009, p. 285) concludes, "is worth dividing up into its smallest possible components".
One of the defining characters of a MOOC is that it is supposed to be massive which is in line with the accepted estimate that around five thousand people constitute the optimum size of a user-base for crowdsourcing purposes. Yet, drawing on "Sturgeon's Law" (Principle 7: Remember Sturgeon's Law), which stipulates that "90 percent of everything is crap", Howe warns us that "real talent will remain a rare commodity" and that crowdsourcing's main merit lies in "providing a previously non-existent outlet for this talent" (Howe, 2009, p. 286). OT12 was - to the best of the project team's knowledge - the first endeavour of its kind.
One important purpose of OT12 was to evaluate whether crowdsourcing via a MOOC could be used to translate OER, and for that purpose we selected Learning to Learn, an OER from the OUUK as one of the resources to be translated. The resource teaches basic study skills to adult students returning to education. It is part of the Bridge to Success (B2S) project, a US-based project funded by the Bill and Melinda Gates Foundation and the William and Flora Hewlett Foundation to encourage student retention and progression in college, especially amongst disadvantaged students or those that have little previous experience of education at this level. The team behind Learning to Learn was keen to have this OER translated into Spanish hoping that this would encourage learners from the sizeable Hispanic community in the US to take up and stay in further or higher education. Thus, rather than looking for the perfect translation ("real talent") they welcomed the idea that crowdsourcing the translation would make the material accessible to many more users, even though that meant that the resulting translations might not necessarily qualify as "professional".
Yet, at the same time Howe's Principle 8 (Remember the 10 percent, the antidote to Sturgeon's Law) also applies to OT12. When asked in the pre-course survey how they thought they'd be able to help their fellow learners, 25% of participants replied that they had either academic knowledge or practical and professional experience of translation, and that they would be willing to share this with others. There was evidently a high level of translation expertise in the group, and this largely corroborates Downes' (2012) observation that MOOC participants tend to come prepared.
OT12, MOOCs and Communities
We understand MOOCs as events (Cormier, 2010) or, following the principles of connectivism, catalysts for starting conversations within a network (Downes, 2011), and therefore felt that a MOOC might be a suitable way of engaging online communities of translators, language teachers and learners, and those interested in OER, in the crowdsourcing of translations for OER.
Our MOOC proved that it was possible to bring together a community interested in finding out more about open translation and to use the MOOC as a way of widening the reach of that community. Balch (2013) hypothesizes that people are more likely to enrol in a MOOC than in a comparable university course, simply because there is no cost associated and no changes required to their lifestyle. Equally, they are more likely to withdraw from the course or to follow it only partially precisely because their economic and personal investment is so much lower. This is reflected in our MOOC, where the level of involvement from each of those who registered their interest varies considerably from those who participated actively throughout, to those who signed up and received the biweekly digest but did not take part in the activities (it is interesting that only one request to unsubscribe from the digests was received from the 600 people in the mailing list). The level of participant investment varied considerably and participants chose their own paths, more or less visibly, through the MOOC.
Whilst the MOOC did, to some extent, fulfil its role of bringing a community together, effective collaboration was not consistently achieved so the outcomes in terms of translation output were variable. Collaboration in the Portuguese team was particularly effective due to the high digital literacy skills shown by some of the members of the Portuguese community and the leading role taken by a couple of its most active members ("Hi everyone on the Portuguese team (…) so happy to see that we have finished the translation! Well done everyone!"). In other teams, particularly those with fewer members, collaboration was less smooth and fruitful.
However, in the evaluation of the MOOC, participants commented on how the course had helped them discover volunteer translating projects that already use crowdsourcing of translation and, in some cases, open tools, which some had joined or were thinking of joining ('I was impressed by the idea of Global Voices, and I plan to investigate that further'); and how they had gained an understanding about OER, Creative Commons, and the open translation 'movement' ('[it was surprising] that this sort of thing was going on in such a wide scale.'). In that sense, a MOOC such as OT12 might have a role to play in bringing together professional or new translators so they can find out about the opportunities to volunteer as translators of open resources. Indeed, since the end of the MOOC, we have used the mailing list to contact the community about open translation opportunities such as the video translation challenge announced by open.michigan at the University of Michigan, US.
This paper has argued that translation and localization are important to increase the portability, visibility and reach of OER; and that crowdsourcing could be the only feasible, affordable solution to the huge challenge of translating open educational content into the many languages that potential users speak. However, there are challenges in terms of quality assurance and in establishing the scope of the translation and its suitability for different contexts.
We have provided some insights into how the world of open education can harness existing open translation models, and learning based on the use of social media, such as happens in MOOCs, to further the openness agenda. MOOCs can be used as tools to raise awareness about open translation and the need to transcend linguistic barriers, but ultimately systems need to be put in place to enable the work of volunteer communities and help them prioritise translations according to different criteria: need, preference, popularity, etc. A translation hub for all OER, along the lines of some of the well-established crowd-sourced large-scale translation projects (Wikipedia, TED Talks, Global Voices), would help achieve such a result.
The OT12 MOOC was partly funded by a Teaching Development Grant from the Higher Education Academy.
Balch, T (2013). About MOOC Completion Rates: The Importance of Student Investment. Available from http://www.augmentedtrader.wordpress.com/2013/01/06/about-mooc-completion-rates-the-importance-of-investment [Accessed 9 July 2013]
Benkler, Y (2006). The Wealth of Networks. New Haven and London: University Press.
Cormier, D (2010). What is a MOOC? At http://www.youtube.com/watch?v=eW3gMGqcZQc [Accessed 1 October 2013]
Downes, S (2011). 'Connectivism' and Connective Knowledge', Huffpost Education. Available from http://www.huffingtonpost.com/stephen-downes/connectivism-and-connecti_b_804653.html [Accessed 9 July 2013]
Downes, S (2012). 'What a MOOC does'. Available from http://www.downes.ca/post/57728[Accessed 9 July 2013]
Graddol, D (2000). 'The Future of English?' The British Council. Available from http://www.teachingenglish.org.uk/sites/teacheng/files/learning-elt-future.pdf [Accessed 9 July 2013]
Howe, J (2009). Crowdsourcing. Why the Power of the Crowd is Driving the Future of Business. New York: Random House
Hyde, A (2009). Open Translation, in FLOSS Manual in Open Translation Tools. Available from http://en.flossmanuals.net/open-translation-tools/ [Accessed 9 July 2013]
Lane, A B, McAndrew, P and Santos, A (2009). The networking effects of OER. Available from http://www.ou.nl/docs/campagnes/icde2009/papers/final_paper_279lane.pdf [Accessed 9 July 2013]
Lane, L (2012). 'Three Kinds of MOOCs'. At http://lisahistory.net/wordpress/2012/08/three-kinds-of-moocs/ [Accessed 9 July 2013]
Lee, M, Lin, M and Bonk, C (2007). 'OOPS, Turning MIT Opencourseware into Chinese: An analysis of a community of practice of global translators', IRRODL, Vol 8, No 3. Available from http://www.irrodl.org/index.php/irrodl/article/view/463/980 [Accessed 9 July 2013]
Ludewig Omollo, K (2013). Help us translate educational videos about microbiology and disaster management from Michigan, Ghana, and East Africa. Available from http://www.open.umich.edu/blog/2013/01/28/help-us-translate-health-oer-videos/ [Accessed 9 July 2013]
OLnet (2009). 'What are the barriers to reusing/remixing OER?' Available from http://www.olnet.org/node/119 [Accessed 9 July 2013]
TESSA (2012). TESSA, Teacher Education in Sub-Saharan Africa. Available from http://www.tessafrica.net/files/tessafrica/Briefing_note_general_June_2012.pdf [Accessed 9 July 2013]
Wolfenden, F, Buckler, A and Keraro, F (2012). OER Adaptation and Reuse across cultural contexts in Sub Saharan Africa: Lessons from TESSA (Teacher Education in Sub Saharan Africa). Journal of Interactive Media in Education. Available from http://oro.open.ac.uk/31439/ [Accessed 9 July 2013]
 Personal communication with Tim Seal, TESS-India Technical Director.