Translation and Participatory Media:

Experiences from Global Voices

ranslation has traditionally occupied a background position in international media despite its pivotal role in facilitating the flow of global news (Bani 2006). In large part this is because translation is subsumed in news agencies within the journalistic tasks of writing and editing; many of those engaged in news translation do not even see themselves as translators, but rather as journalists (Bassnett 2006; Bielsa 2007). Integrated into other duties and overshadowed by more high-profile functions, translation is invisible in international media as it is elsewhere (Venuti 1995), but in a way that is arguably more pervasive.

Background

From its inception, Global Voices has been an experiment in new media. A meeting in late 2004 at Harvard University's Berkman Center for Internet and Society, held at a time when blogging was just taking off in many regions of the world, was the starting point for the project, which has since grown steadily in size and scope (MacKinnon 2004). Ethan Zuckerman, one of two co-founders, had shown through his own research that GDP attracts mainstream media attention to a country more than any other single factor, with consequences that are damaging in areas such as trade and international assistance (Zuckerman 2003). The advent of participatory media, in particular blogging, held hope that this trend could be reversed.

The question was, however: how do you make something like "news" out of the personal blog entries of people living in a faraway part of the world? Through funding from a variety of sponsors, the project that was born set out to answer this question, providing daily summaries of "conversations" (blog posts, comments, forum discussions, etc.) complete with the context necessary to understand region-specific issues. The project's slogan, "The World is Talking, Are You Listening?," embodies its underlying focus on listening, one that differentiates it from the traditional reporting style of mainstream media. It helps in understanding the current operations of the organization, and in particular the way it has dealt with language and translation, to keep this background in mind.


Organizational composition

Global Voices is unique in terms of its organizational composition, borrowing editorial elements from traditional media and yet very different in structure from a typical top-down institution. Oversight, editing, and technical support are coordinated through a team of roughly twenty people, who apart from one full-time managing director all contribute on a part-time basis, and are compensated for their work. Aside from this resemblance in editorial roles, the project is vastly different from mainstream media in most respects, notably: (1) it is entirely virtual, with coordination carried out online through mailing lists, IRC chat, and an internal blog; (2) it incorporates as a major part of its operations a much larger group of volunteers (typically about 80 people); and (3) editors perform a variety of roles beyond actual editing, acting as coordinators, information gatherers and translators.

The following is a rough breakdown of key positions involved in article-writing within the project:

• Regional Editors.  Responsible for writing posts that link to, contextualize and translate the writing of bloggers in one of a number of different regions (Caribbean, Caucasus, Central Asia, Eastern & Central Europe, Latin America, Middle East/North Africa, North East Asia, South East Asia, South Asia, Sub-Saharan Africa).  Regional Editors also co-ordinate volunteer contributors (authors) for their region. As mentioned in the introduction, North America and Western Europe are for the most part excluded from the regions that are covered.
• Language Editors. Responsible for writing posts that link to, contextualize and translate the writing of bloggers in one of a number of different languages (Arabic, Chinese, French, Persian, Portuguese, Russian/Belarusian/Ukranian, Spanish, Japanese, Korean). Language Editors also co-ordinate volunteer contributors who translate posts for the same language. Note that there is overlap between regions covered by Regional Editors and Language Editors.
• Authors. Volunteer contributors (typically bloggers) who summarize blog posts in a region, language or topic with which they are familiar. In total the group of authors numbers around 80 people. (Note that the title "author" is somewhat misleading in that while authors write articles, these articles describe blog conversations by other bloggers and generally do not present original views of the author.)

In addition to the positions above, there are two co-founders, one managing director, one managing editor, two members of a technical and design team, as well as editors for specific topics (environment, podcasts, videos) and directors for advocacy, outreach, and translation.

Global Voices also has a board of directors made up of seven members, four representing the organization's founders, contract employees and volunteers, and three external advisers.


Languages and translation

The role of translation within Global Voices can be difficult to distinguish due to the degree to which it is embedded within the project at an implicit level. The task of Language Editors, for instance, is to translate and contextualize blogs written in a particular language, the term "language" relating to the task only at a very general level. Similarly, Regional Editors perform translation when incorporating non-English language blogs in their coverage, but the translation component of their role is not often mentioned. The Global Voices Manifesto frames the goals of the project in terms of the right to free speech and the right of all voices to be heard, with language and translation again implicit.²

The place that translation occupies within Global Voices arises not by design, but through the history of events that brought the project to its current form. At the very start of the project, a group of regional editors was hired to introduce to a global audience the writing of bloggers in different regions of the world (specifically the non-Western world). It was hoped at the time that "bridgebloggers" (Zuckerman 2007), local bloggers who write about a particular region in English (hence forming a "bridge" with the wider world), would be sufficiently numerous to fulfill this role. More so than translation, also considered a critical issue at this early stage (MacKinnon 2005), the focus was as such on context: the historical and cultural background required to make a foreign (English-language) blog entry understandable to an outsider.

It was soon realized, however, that there were not nearly as many bridgebloggers as would be needed to make this kind of coverage work on its own. It was at this point that Global Voices, through funding from Reuters and other sponsors, began hiring translators to cover particular languages ("lingospheres"). The current organization, wherein there is overlapping coverage of regions and languages, was in this way born.³

This development of the project, from a focus that started as region-based to one that is increasingly linguistic, parallels a transformation that some have predicted will affect the Internet as a whole. Clay Shirky, one of the most well-known commentators on web culture, wrote nearly ten years ago that in the next century, "the definition of proximity [will change] from geographic to linguistic: two countries [will] border one another if and only if they have a language they can use in common" (Shirky 1999). Global Voices, as an organization, interweaves both geographic and linguistic spheres into the same citizen media project, and as such can be seen as reflecting this underlying trend of a shifting proximity.

Just as linguistic proximity grows in importance, regional proximity nonetheless remains critical in grounding events to local realities. To this end, whereas traditional media organizations have typically "parachuted" foreign nationals into a country, Regional Editors at Global Voices are generally local to the region, as are most Language Editors. This is significant in a number of ways: editors speak a local language, are familiar with regional social and political issues, and are embedded within local social networks. It likewise enables Global Voices to keep a constant eye on countries in regions that the mainstream media covers at best sporadically.

While this regional connection is very important in finding and transmitting stories that would not otherwise receive media attention (the raison d'être of the project), it poses challenges in other respects. In particular, much of the writing at Global Voices is performed by authors and editors for whom English is a second language, a practice that runs counter to the standard philosophy of translating only into one's mother tongue (Newmark 1988). For a project that attempts to incorporate local voices into a global conversation, this type of second-language English writing is a practical reality that cannot be avoided. It can be nonetheless problematic when the English-language text must itself be translated, a situation that arises within Project Lingua, described in the following section.

If the role of languages and translation within Global Voices is difficult for the newcomer to understand, then Project Lingua ("Lingua" for short), a community that translates Global Voices articles from English into (at last count) 14 different languages, presents a yet more complicated picture. Composed of teams each responsible for a single language, Lingua has as its goal to bring Global Voices content to new linguistic audiences. It accomplishes this goal through daily translations of English-language content from the main website.⁴

History of the project

While the basic function of each Lingua team is easy to state (translate English-language articles into a given language), the organizational relation of the project as a whole to Global Voices is complex and sparsely documented. In large part the latter is due to its very short history, Lingua having been launched as a project at the end of 2006. Tracing this history of the project, which has grown from a small group of translators into what is today a community with an estimated 50 active members, goes some way to explaining the role it has taken on within Global Voices.

It is important to note, firstly, that the present organization of Project Lingua was never designed in advance. Unlike Global Voices itself, whose makeup was determined through various meetings and online discussions, Lingua was an initiative that emerged from the community of Global Voices readers, and as such it developed in a very different way.

Portnoy Zheng, a Taiwanese blogger and current director of Project Lingua, began in September 2005 to translate (pro bono) selected Global Voices articles, which he posted at his own blog. At the time a student of journalism and communications, Zheng "fell in love with [Global Voices] at first sight" and felt compelled to bring the site's content across the Chinese language barrier to an audience of Taiwanese netizens. Like many others who have joined the project since, his choice to translate Global Voices content was motivated by a dearth of international news in his own language community (Góes 2008).

This approach of translating selected full-length articles into other languages, referred to as the "Portnoy model," developed completely independent of Global Voices itself. The founders of the project, Ethan Zuckerman and Rebecca MacKinnon, had never considered whether such a translation model was desirable, because they had never considered that it was even possible.⁵ When MacKinnon, a Chinese-speaker herself, discovered Zheng's work, she thus naturally encouraged him to continue. No legal issues arose as all articles at Global Voices are available for use under a Creative Commons license.⁶

Discussions at the second Global Voices Summit in December 2006 made clear that while Zheng had been the first to actually translate Global Voices content, he was not alone in his thinking; many within the community, it was discovered, were eager to see the site translated into their native languages. Project Lingua was the name given to the project that emerged from these discussions: a cluster of satellite translation teams, each allocated a sub-domain of the main site, running a separate localized installation of the same blogging software (WordPress).

Project Lingua has grown considerably since 2006, at present numbering several dozen active contributors and covering 14 different languages. Whereas Zheng's original translations were the work of one enthusiastic fan, the project now comprises a major component of the Global Voices project. Its unique history, and the position it plays in transmitting content to fundamentally different linguistic audiences, sets it however apart from other sub-projects.⁷ The latter explains to some degree the largely background role Lingua continues to play, despite its expanding size and reach, within Global Voices as a whole.


Organizational composition of translation teams

Project Lingua currently comprises language teams covering 14 different languages: German, Spanish, French, Malagasy, Portuguese, Albanian, Macedonian, Arabic, Farsi, Bangla, Hindi, Chinese (Traditional and Simplified), Japanese, and Italian; one language (Russian) was once active but has since been discontinued for lack of contributors. Each team is made up of one editor and a group of volunteer translators, the latter varying in size and level of activity. While participation in the project fluctuates rapidly, at the time of writing this article the teams with the largest number of volunteers were Spanish and Chinese. Spanish, Malagasy and Bangla teams produced the highest number of translations over the past six months.⁸

Lingua is highly virtual, with contributors regionally dispersed even within individual language teams. Members of the Spanish team, for example, are based in Peru, Argentina, Venezuela, Spain and the US., while the Arabic team includes members from Syria, Lebanon, Egypt, Iraq and Bahrain. Lingua is also very diverse in terms of the backgrounds of people involved in the project. Common profiles include blogger, journalist, translator, and student, but there are also lawyers, system engineers, professors, IT professionals, financial analysts and web designers.

One of the encouraging developments in Lingua is growth in teams representing smaller and more receptive niche communities, the availability of global news in such languages being relatively scarce. The Malagasy team, in particular, is strongly motivated by preservation of language, the majority of blogs in Madagascar being written in French despite the fact that the country is majority Malagasy-speaking. Whereas many Lingua teams have struggled to attract attention in their local language community, the Lingua Malagasy site is rated as one of the top ten Malagasy blogs.⁹


Communication and co-ordination

While diversity of region and background is common across the entire Global Voices community, the nature of Lingua as a decentralized multilingual project poses significant challenges in terms of co-ordination and communication. In particular, while Lingua has a director of operations (currently Portnoy Zheng, formerly one-time Francophonia Editor Alice Backer), there is no editorial oversight covering the entire project. This situation is inevitable considering the range of languages involved, but it nonetheless conflicts with a policy of maintaining editorial standards across all Global Voices content. Constraints on non-English content have thus been imposed on translation teams to maintain consistency across all languages (see Discussion).

There is a deeper challenge, moreover, of a context gap between Lingua and the rest of the Global Voices project. As an organization, Global Voices is held together by unifying elements: a general level of fluency in English, common points of reference (blogging, new media, activism), and the existence of a core team and board of directors with a broadly shared vision. These elements do not exist uniformly across Lingua, which is more strongly tied together by the language-specific function of translation.¹⁰ While translators must possess a general English reading ability in order to contribute, this does not imply a familiarity with English as an operational day-to-day language of communication; nor does it guarantee knowledge of the tools of blogging, an integral part of the project as a whole.

For these reasons, Lingua translators are understandably less active in general mailing list discussions than are Global Voices authors and editors. Likewise, limited knowledge of blogging tools in many cases demands that they receive special training. Separate mailing lists, websites, and even tutorials have accordingly been created by translation teams, in their respective languages, for new volunteers. These systems are essential in the day-to-day operations of these teams, but they also create a communication gap, both internally between translation teams and with respect to the rest of the Global Voices organization. Finding a way to encourage cooperation between Global Voices contributors and Lingua translators, which would help streamline the translation process, is as such viewed as an important target in assuring future success of the project.

Translation flow in Global Voices and Project Lingua

Global Voices describes itself as an organization that "aggregates, curates, and amplifies the global conversation online." Having outlined the role of languages and translation both in Global Voices as a whole and in Project Lingua in particular, we are now in a position to state more clearly how translation fits into this process.

The Global Voices component of this process may be thought of in three steps:

1. Search. Regional Editors, Language Editors and Authors search for topics being talked about in blogs and other user-generated content in their region or language.
2. Select. Once an issue or topic has been found, entries and background information on this topic are selected. Reference information (linked to but not generally quoted) may include newspaper articles and other sources that are not strictly user-generated content.
3. Compile. Selected passages from blog entries and other user-generated content are compiled into an article (or series of articles), with background information and reference links incorporated to provide context for a global audience.

Note that this is simply a schematic for the process, and different contributors carry out this task in different ways.

Where source documents are in a language other than English, translation is performed either at step 2 or step 3. In cases where video content is included, subtitle translation is also sometimes required. This is generally carried out using a free online service called dotSub (see below).

Once an article has been created and posted at the main Global Voices page, it is available for Lingua teams to translate. Volunteers are free to select whichever article they wish to translate, as long as it has not yet been taken by someone else in their language team. When the translation is completed and posted, a link appears at the head of the original (English-language) article pointing to the translated text on the corresponding Lingua site.

As noted earlier, procedures for co-ordinating and reviewing translations vary considerably across Lingua teams. The following are a few practices that are employed:

• Mailing list + editor review. Translators announce which article they plan to translate through a mailing list, and one or more editors proofread all translations prior to publication. This is the most common approach.
• Wiki + peer review. In this system, used by the Chinese team, each article is proofread by another team member, sent back to the translator for confirmation, tagged as "ready for publication," and finally published. Everything up until the last stage is carried out on a separate Wiki site used within the team for translating, editing and proofreading, with a mailing list used for additional discussions.
• Direct publication. Some groups also allow certain translators to directly publish translations (without review) in cases where their translation quality has already been tested and they are familiar with the blogging software. Co-ordination again generally takes place through mailing lists.

Collectively, the procedures outlined above summarize the process of aggregation and amplification within Global Voices and Project Lingua. From beginning to end, this process carries a set of source texts in participatory media across barriers of region and language to an audience which, in most cases, it would never have otherwise had. Despite the complexity of the underlying procedures, it is this function of amplifying individual voices that distinguishes the project, and from this process that the project draws its support. Acting as a linguistic bridge, translation plays an integral role in this process.

Originality and the lost context problem

One of the key issues facing Project Lingua at the present time concerns the publication of original content in non-English languages. From the start, Global Voices management made it clear that such content could not be published on Lingua sites unless it already had an English equivalent at the main site. The underlying issue is one of quality control: content with no English version cannot be validated, as there is no equivalent in Lingua teams of a Regional Editor to edit and proofread the original writing of volunteers. In ruling out such original non-English writing, the prototype Lingua site has been constrained to act as a mirror of Global Voices, assuring that translations, while not readable, are at least in line with main site content.

While necessary from the viewpoint of management, this constraint poses a problem for Lingua teams attempting to promote Global Voices content to non-English audiences. The issue that arises might be termed the "lost context problem": separation of an original text (Global Voices article) from the context within which it was originally embedded. While loss of context occurs in any translation, the problem is particularly acute in this case as Global Voices articles are themselves assembled from foreign (often translated) content embedded in contextualization (the "compile" step described above).

Loss of context manifests itself in a number of different ways:

• Background knowledge. Articles assume that readers have background knowledge in areas that non-English audiences may not be familiar with. An article about Mauritania mentioning genital excision, for example, posed a challenge for a Malagasy translator since the concept is foreign to audiences in Madagascar (Andriamananjara 2007). Translator notes can to some degree help in solving this problem, but there is a limitation on such notes imposed by the constraint on original content.
• Links and references. Where background is required, Global Voices articles commonly link to English-language websites and blogs, but these become inaccessible once the article is translated. Links to English-language references are thus sometimes replaced with links to non-English equivalents where such texts exist (for example in the case of Wikipedia entries). Beyond this scope, however, there is no possibility to link to blogs in the language that the article is being translated into. This prevents Lingua teams from connecting with their local blogosphere, a crucial aspect of any blogging project.
• Subject. Topics taken up by Global Voices may implicitly assume a social context that, while natural for an English-speaking audience, may not be appropriate for readers of a different language. Audiences in South East Asia, for example, are more reluctant to engage in themes such as politics and human rights, and it has been suggested that such themes (common on Global Voices) may detract from readers contributing to the project. Translators can simply choose not to translate such posts for their audiences, but this again is at best a partial solution.
• Presentation. The writing style and presentation of Global Voices may evoke an unintended response among a different linguistic audience. One editor noted that to francophone audiences in France and Africa who read websites like Global Voices, the style of articles may feel foreign. The attitude of "let's unite to make this right," for example, appropriate as a means of engaging and motivating a U.S. audience, can appear as overly optimistic if translated without adaptation into French.¹¹

All of these issues may be seen as common to any localization project. Unlike comparable commercial websites, however, Global Voices did not set out to be a multilingual project; the expansion into different language communities developed rather as an initiative of the community. The organization can as such be seen as having had to cope with an extremely challenging and unanticipated problem (localizing a global media project into over a dozen languages) with scarce resources. Lost context is central to this problem in that it not only raises questions regarding the limits of Lingua as a translation project, but also highlights the nature of Global Voices as an English-centric organization.

A number of solutions have been proposed for the lost context problem. One that has attracted attention would allow for articles to be posted in a language other than English, at the respective Lingua website, on condition that an English translation existed beforehand. One of the benefits of this approach is that it would allow bloggers who are not fluent in English to participate in the project, while also satisfying the need for editorial oversight.¹² In itself, however, this would not solve the problem of lost context in cases where foreign content is translated through English.


Context and news translation

The problem of lost context can be related to the process of textual transformation which occurs in the translation of international news. Analyzing English-Finnish news translation, Hursti observes that in news production translation texts undergo reorganization, deletions, additions, and substitutions. Most reorganizations, he observes, are not due to differences in language, but rather motivated by conscious decisions to refocus the target text to fit the receiving audience (Hursti 2001).

More recently, Bielsa examined frequent modifications applied to source texts in news translation, including change of the title or lead, elimination of information redundant to a particular audience, and addition of background information. One criteria used to justify such alterations is the background knowledge of the audience, mentioned as a key source of lost context in the last section. Bielsa observes that the effect of modifications make the translation "more like an original, new text, specifically suited to the needs of the publication in which it appears and the readers to which it is targeted" (Bielsa 2007:142-143).

It is interesting to note that whereas modifications of the kind listed above are integral to the procedure of summarizing blog conversations in Global Voices (the "select" and "compile" steps mentioned above), they are absent from the translation step carried out by Lingua translators. The dividing line between "journalist" and "translator," blurred in news translation through the elevation of the former over the latter, may as such be seen as manifesting itself at an organizational level in the different functions carried out by Global Voices and by Lingua.

Inter Press Service (IPS), a global news organization which has a similar policy of "giving a voice to the voiceless," employs a different approach in which context is added to translated stories to "balance what is local and what is global." Miren Gutiérrez, editor-in-chief of IPS, notes for example that while the concept of "paramilitary" is all to familiar in Latin America, it is not so elsewhere: "the entire meaning and impact of the article might have been lost on readers in the English-speaking world, even though enough information was available for the average Spanish-speaker" (Gutiérrez 2006:31). A closely-related problem confronts Lingua teams in the translation of Global Voices articles, where source texts are implicitly contextualized for an English-speaking audience. The approach of IPS and other multilingual news organizations may as such serve as a useful model in broadening the scope of translation within Global Voices.

Relation to fan translation

The lost context problem also indicates a connection with another form of translation, referred to as "fan translation." Díaz Cintas and Muñoz Sánchez studied practices in the volunteer-based translation of Japanese anime subtitles ("fansubs"), one of the most prevalent forms of fan translation (Díaz Cintas and Muñoz Sánchez 2006). While fan subbing dates back to the 1980s, the activity took off in the 1990s with the arrival of cheap computer software and tools.

There are a number of similarities that link translation within Global Voices to the translation of fansubs. Unlike news translation, fan translation of subtitles is a volunteer-based initiative, closer in this sense to the motivations of Global Voices and Lingua contributors. Also like Global Voices, English is used in fansubs as a "pivot language" to translate Japanese into other languages, and most fan translators are not native-English speakers. More broadly, Díaz Cintas and Muñoz Sánchez note that fan subbing "lies at the margins of market imperatives and is far less dogmatic and more creative and individualistic" than traditional media translation (Díaz Cintas and Muñoz Sánchez 2006:51); as a nonprofit project focused on the translation of user-generated content, Lingua could be described in much the same way.

The lost context problem manifests itself in fan subbing and is dealt with using an approach that is different from professional subtitle translation, and also different from news translation. Díaz Cintas and Muñoz Sánchez note that in fan subs, "certain cultural referents such as the name of places, traditions and other celebrations are explained by using translator's notes and glosses." (Díaz Cintas and Muñoz Sánchez 2006:46) Translation notes or comments are also added by some fansubbers as additional background information, targeted at an audience of people interested in the world of anime.

In relation to the problems of context outlined above, it has been noted in discussions that the audience for Global Voices content is not well-defined. Resolving the lost context problem may as such boil down to identifying local audiences and providing relevant contextual notes aimed specifically at those audiences. One group that has shown interest in Global Voices content, and particularly in Lingua translations, are members of diasporas.¹³ Borrowing from the fansub model, a concrete step in overcoming lost context would involve supplementing Lingua translations with additional information relevant to such readers. A similar approach could be applied to many other interest groups.

Both news translation and fan translation generally involve content that is copyrighted. Translation of content that is free to use (usually under a Creative Commons or GPL license), termed "open translation," differs in that there are generally no issues of legality or ethics.¹⁴ All Global Voices content is licensed in this way.

The number of open-translation based community translation projects has increased in recent years. The following is a short list of examples:

• Cucumis (www.cucumis.org): a volunteer translation community with contributors from over thirty different languages. Users can offer translations, earning points which they can then use to have their own texts translated. The system is entirely volunteer-run and all translations are performed for free. A handful of Cucumis volunteers recently joined the French translation team at Lingua.
• dotSub (www.dotsub.com): A website that allows users to add timed subtitles in any language to a given video, in such a way that the subtitles can later be translated by other users without reference to the original video. This makes it easy for Lingua translators to incorporate video content into their translations, and the service is often used for translating videos featured in Global Voices articles.
• Worldwide Lexicon (www.worldwidelexicon.org): open-source project oriented toward developing tools that enable users to easily create, edit and share translations online. Translation of user-generated content such as blog entries is a major focus of the project.

The translation of open-source software and documentation by volunteer translators is another prominent example of open translation with a long history (Sasaki 2007). More recently, the use of multilingual wiki engines for collaborative online translation has also been investigated (Désilets et al 2006).

The above projects share certain features in common: they are oriented specifically toward translation, to varying degrees are platform-based, and are multilingual without emphasizing one language over another. Global Voices, in contrast, does not focus specifically on translation, is more strongly tied to its community than to its platform, and centers on English as a common language of communication.

The current article has highlighted various problems related to this lack of emphasis on translation, and to the assumption of a context implicitly geared toward an English-speaking audience. It is nonetheless important to reiterate that Global Voices would not be the project it is if its focus was exclusively on translation. The core mission of shining a light on individual voices, more so than translation itself, is what draws people to the project. The goal as such should not be to become a platform for translation, but instead to act as a space within which translation, traditionally invisible in the news, takes on a new and more meaningful role in bridging global perspectives.

With the Internet becoming more multilingual by the day, and international mass media increasingly pulling away from the costly business of foreign correspondents and foreign bureaus, there is a growing need for local voices to fill in the gaps of "global news." As a mediator between languages and cultures, translation plays a key role in this process. Connections between the new and evolving realm of participatory media, and the age-old practice of translation, are however few and far between. In identifying and explaining the role of translation in one of the most influential citizen media projects in the world, it is hoped that the present article may act as a first step bridging these very different areas.

Michael Cronin (Cronin 2003) has written in depth about the connections between translation and globalization, ideas that apply equally well to the landscape of global conversations discussed in this article. His words serve as a closing reminder for those engaged in the creation of participatory media of the significance of translation in a globalized world:

² See: http://www.globalvoicesonline.org/about/gv-manifesto

³ For more background on the role of translation in Global Voices, see: "Language and translation on Global Voices," My heart's in Accra (blog of Ethan Zuckerman), December 16, 2006. http://www.ethanzuckerman.com/blog/2006/12/16/language-and-translation-on-global-voices/

⁴ See: http://www.globalvoicesonline.org/lingua

⁵ Ethan Zuckerman had in fact originally envisioned a different model wherein people created their own separate GV-like sites for different languages (personal communication). Some Lingua editors have mentioned that they view such sites as an ultimate future goal, but none have yet been created.

⁶ For more about Creative Commons licenses, see: Lessig, Lawrence, "The Vision for the Creative Commons: What are We and Where are We Headed? Free Culture," in Brian Fitzgerald (ed.), Open Licensing: Cultivating the Creative Commons, Sydney: Sydney University Press, 2007. http://ses.library.usyd.edu.au/handle/2123/1572

⁷ There are three other sub-projects within Global Voices: Advocacy (http://advocacy.globalvoicesonline.org), Rising Voices ( http://rising.globalvoicesonline.org), and Voices without Votes ( http://voiceswithoutvotes.org).

⁸ See: http://globalvoicesonline.org/lingua/posting-stats

⁹ Lova Rakotomalala (editor of GV Malagasy), personal communication.

¹⁰ Note however that in fact many Lingua editors are also GV authors, and form a bridge between Global Voices and the larger group of volunteer translators.

¹¹ Claire Ulrich (editor of GV French) notes in an email (June 9, 2008): "GV and other US communities often have a 'can do', 'let's unite to make this right' and 'we are great!' attitude that is part of the American way to communicate, engage, motivate. It sounds over optimistic and boyscoutish to the francophone readers."

¹² Some teams have already experimented with this approach. See for example: Portnoy Zheng, "Taiwan: An Ironic Human Rights Day," Global Voices Online, December 24, 2007. http://www.globalvoicesonline.org/2007/12/24/taiwan-an-ironic-human-rights-day/

¹³ Claire Ulrich (editor of GV French), personal communication.

¹⁴ The first Open Translation conference was held in Zagreb, Croatia in late November, 2007. See also: "Voices from Open Translation Tools 2007": http://blip.tv/file/522086/

Andriamananjara, Mialy (2007). "Mauritania: Ignorance and Tradition," Global Voices Online, August 30, 2007. http://www.globalvoicesonline.org/2007/08/30/mauritania-ignorance-and-tradition

Bani, Sara (2006). "An analysis of press translation process," in Proceedings of the Translation in global news conference, University of Warwick, 35-45.

Bassnett, Susan (2006). Introduction to Proceedings of the Translation in global news conference, University of Warwick, 5-7.

Bielsa, Esperança (2007). "Translation in global news agencies," Target 19 (1), 135-155.

Cronin, Michael. Translation and Globalization. Routledge: London, 2003.

Désilets, Alain et al. (2006), "Translation the Wiki Way", in Proceedings of the WIKISym 2006 - The Conference Wiki of the 2006 International Symposium, Odense, Denmark, August 21-23, 2006. http://iit-iti.nrc-cnrc.gc.ca/publications/nrc-48736_e.html

_____________ (2003). "Global Attention Profiles - A Working Paper: First Steps Toward a Quantitative Approach to the Study of Media Attention," The Berkman Center for Internet and Society, Research Publication No. 2003-06, August 2003. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=487943