Media Monitoring Technologies in International Media ...

This article was downloaded by: [Heidi A. Campbell] On: 14 March 2013, At: 10:04 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Broadcasting & Electronic Media Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/hbem20

Language and the Boundaries of Research: Media Monitoring Technologies in International Media Research a

a

Randolph Kluver , Heidi A. Campbell & Stephen Balfour a

a

Texas A&M University

To cite this article: Randolph Kluver , Heidi A. Campbell & Stephen Balfour (2013): Language and the Boundaries of Research: Media Monitoring Technologies in International Media Research, Journal of Broadcasting & Electronic Media, 57:1, 4-19 To link to this article: http://dx.doi.org/10.1080/08838151.2012.761701

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/termsand-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation that the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages

Downloaded by [Heidi A. Campbell] at 10:04 14 March 2013

whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

Language and the Boundaries of Research: Media Monitoring Technologies in International Media Research Downloaded by [Heidi A. Campbell] at 10:04 14 March 2013

Randolph Kluver, Heidi A. Campbell, and Stephen Balfour This article argues that data-driven research in global media has not kept up with the inherent interest in internationalization due to a number of factors, including access to global media and the lack of foreign language skills among many media researchers. Thus, the field still remains focused on analysis of Western media, leading to misunderstanding the geopolitical impact of global media. The article explores emerging media monitoring technologies which have arisen out of developments in natural language processing and machine generated translation. Finally, the article explores innovative research programs that utilize these technologies for global media research and education. Media studies as a field exists in a rapidly globalizing world in which media is one of the key carriers of the very processes of globalization (Appadurai, 1996). Decades after the famous ‘‘MacBride Report’’ which decried the lack of news flow from the developing world to the developed, global media studies still remains largely focused on analysis of Western media with very little attention paid to important global media sources (MacBride, 1980). This is not due to a lack of interest of media scholars; the field actually has a tremendous interest in internationalization (Ekecrantz, 2009). However, an overwhelming amount of media studies research remains rooted in analysis of Western media; the field hasn’t generated a corresponding amount of research on global communication organizations. This lack of attention on global media contributes to what Chakrabarty has called ‘‘assymetric ignorance’’ of news coverage itself, in which critical global events, conversations, and issues remain Randolph Kluver (Ph.D., University of Southern California) is an associate professor of Communication and Executive Director of Global Partnerships at Texas A&M University. His research interests include the political impact and geopolitics of new media. Heidi A. Campbell (Ph.D., University of Edinburgh) is an associate professor of Communication at Texas A&M University. Her research interests include new media, religion and digital culture. Stephen Balfour (Ph.D., Texas A&M University) is an instructional associate professor of Psychology and Director of Information Technology for Liberal Arts at Texas A&M University. His research interests include teaching and learning with technology and the roles of writing and peer review in learning. The authors would like to acknowledge the Technology Support Working Group for access to the technologies presented in this article and for funding that has enabled this research. © 2013 Broadcast Education Association DOI: 10.1080/08838151.2012.761701

Journal of Broadcasting & Electronic Media 57(1), 2013, pp. 4–19 ISSN: 0883-8151 print/1550-6878 online

4


Kluver, Campbell, and Balfour/LANGUAGE AND THE BOUNDARIES OF RESEARCH 5

largely out of the view of the general populace of the developed world (Chakrabarty, 2007). When the experiences, policies, and impacts of a few (Western) media outlets are the almost exclusive focus of scholarly attention, there is little doubt that the findings and theories emerging from that research become the de facto explanatory framework for all of global media (Curran & Park, 2007; Thussu, 2009). The purpose of this article is to examine the impact of this unbalanced academic attention, and to propose the development of methodologies that draw upon advances in digital media to begin to address it. Our argument will be that new media technologies offer unprecedented abilities to begin to address these gaps, and to develop research methodologies that take into account the media experiences of most of the globe. Within the pages of the Journal of Broadcasting and Electronic Media itself, there has been little attention to global broadcast outlets outside of the ‘‘duopoly’’ of US and UK based outlets. A recent search (in July of 2012) of the journal archives (1957–2012) reveals 125 references to CNN, while a similar search for Fox reveals 246 articles. The BBC has been mentioned 117 times. Interestingly, the international version of CNN, CNN International, has only appeared in three articles. In contrast, fewer than a dozen articles have been published on major global media outlets, including CCTV, Al Jazeera, Al Arabiya, and Phoenix, which collectively represent the key media outlets for almost two billion people, or one third of the global population. Likewise, a similar analysis of the Journal of Communication revealed 8 mentions of Al Jazeera, 3 of CCTV, 102 on CNN, and 197 for BBC (Table 1). This is not to say that large US or UK based broadcasters are unimportant; indeed, as Volkmer argues, the launch of CNN International in 1985 ‘‘brought about new forms of global political communication within only a period of only a few years’’ (Volkmer, 1999). But, it does point to a paucity of research on global media outlets that serve the majority of the human race. Similar comparisons with other media types would likely reveal similar gaps between research on Western outlets and those of much of the rest of the world. It seems hardly necessary to justify why the field needs to conduct more research into global international media outlets, but we would like to point out a few obvious reasons. First, as an area of research, media studies, regardless of the disciplinary background, must ultimately take into account global diversity in studying human experience. Although US and UK based broadcasters, for example, are globally important, they can hardly substitute for media outlets that originate in vastly different political, economic, and cultural contexts. One 2002 Gallup survey, for example, showed that among Al Jazeera viewers in 9 Arab nations, ‘‘the attitudes of Al-Jazeera viewers are consistently different from those of other residents of these countries’’ (Saad, 2002). CNN International, for example, claims a global audience of approximately 200 million households and is distributed in over 200 countries (CNN, 2012). Despite that wide footprint, it is typically only watched by Englishspeaking elites and business professionals, and its global coverage is still largely influenced by North American broadcast priorities (MacKinnon, 2004). This limited access means that it still has a limited global impact, in spite of its global presence.

6

Journal of Broadcasting & Electronic Media/March 2013

Table 1 References to Major Global Broadcasters in JoBEM*


Broadcaster CNN CNNI (CNN International) FOX BBC Sky News Deutsch Welle Al Jazeera Al Arabiya Chinese Central Television (CCTV) Phoenix (China) TeleSUR Univision Telemundo TV Globo

# of references in JoBEM 125 3 246 177 5 12 8 2 9 0 0 11 8 5

*Results obtained by using online archive search of JoBEM at www.tandfonline.com July 12, 2012.

In comparison, the 16 channels offered by CCTV (Central China Television) have a market share of approximately 44% of China’s 1.3 billion persons, while CCTV-9, the English channel, reaches an estimated 85 million viewers in over 100 countries (‘‘The CCTV’s International news channel CCTV-9 is renamed CCTV News,’’ 2010). As Yuan and Webster note, the Chinese television audience has become the world’s largest audience, with CCTV networks holding the lion’s share of that audience (Yuan & Webster, 2006). Likewise, Al Jazeera reaches an estimated 50 million viewers across the Arabic speaking world, and a recent poll has shown Al Jazeera remains the most popular news source across the Arab world, capturing a full 53% of the respondents (‘‘The 2011 Arab Public Opinion Poll,’’ 2011). As Thussu (2009) argues, this means media studies as a field remains highly provincialized, which contributes to an ‘‘epistemological essentialism,’’ in which the theories generated from extensive scholarly analysis remain rooted in the experiences of a small percentage of the global population. Ekecrantz further argues that there is a sort of ‘‘methodological nationalism’’ at work in media studies, in which the methodological (and theoretical) foundations for media studies are developed out of, and in reference to, the dominant institutions of the nation in which media studies scholars are trained (Ekecrantz, 2009). Regardless of the accuracy of this claim, it is true that the majority of academics working in the field are writing and researching primarily for a Western research community and that ‘‘the academic ‘other’ is often added as an afterthought or a token’’ (Thussu, p. 26).



Thus, academic media researchers study the media which is most readily available not just to themselves, but to their academic audience as well, virtually guaranteeing a spiral of silence, so that media outlets from most of the globe never enter public consciousness or the awareness of media scholars. Ultimately, this impacts pedagogy as well. As Downing illustrates, there is little integration of global media into most instruction at top U.S. universities, which he argues is most likely reflective around the globe, thus contributing to the ‘‘continuing construction of this unnerving deficit’’ (Downing, 2009). It is not just media theory and education that suffers from this imbalance of global media, but there are also political and economic consequences, as U.S. policymakers, educators, and citizens remain blissfully unaware of the ways in which Western policies and interests are discussed, disputed, and portrayed in global media. Thus, beyond a vague recognition that Western powers are distrusted in the Arabic, Chinese, or Bahasaspeaking worlds, we have little ability to determine what specifically drives global responses to the policies of the West. As Tehranian argues, global communication can either act as a process for the free and equal exchange of perspective, or it can ‘‘systematically distort perceptions by creating phantom enemies, manufacturing consent for wars of aggression while stereotyping and targeting particular ethnic groups or nations into subhuman categories’’ (Tehranian, 1999). The implications of this are significant, in that it ultimately means universities are abdicating a significant role in helping students learn more about the world around them, which would ultimately enable students to become more discerning and appreciative of the important geopolitical issues which media impact. The problem is not that there are no non-Western media models and theories; indeed, a number of scholars have sought to build theoretical models that don’t overrely on Western experiences (Banerjee, 2009; Sabry, 2009; Zhao, 2009). Moreover, in recent years, there has been significant growth in academic literature examining global players like Al Jazeera, for example (Lynch, 2006; Seib, 2008; Zayani, 2005). The literature is most welcome, but probably reflects heightened geopolitical sensitivities over the impact of one particular network more than it does an awareness of the importance of global media in general. There are probably additional reasons for the lack of research in global media, such as lack of editorial interest or knowledge, the lack of knowledgeable paper reviewers, or even self-censorship, but in spite of the recent work on Al Jazeera specifically, the amount of academic research on global media from China, the Arab world, Russia, and Latin America remains sparse. In spite of this interest in developing global, international, or even transnational, theoretical and conceptual models, it is difficult when there is so little data-driven research on global media outlets to develop these perspectives. So, the problem is not with the intention of researchers, but rather with their access to global media, as there are too few researchers with the access, language skills, and the scholarly background to develop such research. Thus, one of the greatest needs to further the internationalization of media studies is for wider access to global media that enables researchers without the linguistic or cultural background to access that media.

8



The Impact of Digital Media on Global Media Research The impact of digital media technologies on global media has been well documented, both in the popular press and in academic literature. As digital technologies have taken center stage in the production and distribution of content, these processes have become faster and more cost effective, which has presented significant challenges to traditional media corporations and organizations. At the same time, however, it has allowed global media producers to have a greater global footprint and build audiences around the world. All types of media content (e.g., text, video), have benefitted from these rapid and radical changes, and media companies continually experiment with models of distribution that enable them to build audiences, while still maintaining financial solidity. In the early part of the Internet boom, many argued that this would change the underlying dynamics of global communication, and ultimately undermine narrow or provincial perspectives. As global news began to appear on the Internet, there was a promise of an unprecedented ability to interact with, and learn from others around the world. The assumption was that as people from around the world were able to interact there would be a natural recognition of the perspectives from other nations, so the Internet signaled the ‘‘end of nationalism.’’ Soon, however, it became apparent that although the Internet is global, it was not necessarily globalizing. As content on the Web exploded, and the promises of the Internet’s low barrier to entry allowed global news outlets, as well as other types of content producers, from around the world to place content online, it became clear that for technical, linguistic, cultural, and historical reasons, Internet users typically were exposed to content from their own country, rather than from global sources (Best, Chmielewski, & Krueger, 2005; Halavais, 2000; Kluver, 2002). There are multiple explanations for this, including Best’s argument that selective exposure best accounts for choices of online news sources. They found that during the Iraq war, those who disagreed with US engagement in that war sought out global news perspectives in order to confirm their own political perspective, rather than to truly engage with alternative points of view. Specifically, they argued that ‘‘individuals use foreign news sources not simply to satisfy a hunger for political news generally, or even information about foreign events specifically, but primarily to appraise attitude-consistent news stories’’ (Best et al., 2005, p. 65). Halavais examined more technical factors through his link analysis of the World Wide Web, and found that although the Internet is in fact ‘‘more internationalized’’ than other channels of content delivery, national boundaries still have relevance, as networks of hyperlinks tend to remain rooted to one’s home country (Halavais, 2000). Moreover, his study demonstrated certain types of Web sites are more likely to contain global linkages; specifically, scientific and technical Web sites were far more likely to have international links embedded on them than news, sports, or governmental sites (p. 21). US Internet content producers are less likely than those from other nations to link to international sources of information. Perhaps even more importantly, Halavais found that even when Web sites are linked across national



boundaries, there is far less likelihood of linking across linguistic boundaries. This finding lends support to the argument that the World Wide Web doesn’t live up to its potential for creating international linkages. These studies and others indicate the rise of digital media has not, in fact, led to the international exposure that earlier proponents of the Internet, and of global media, expected. Yet it doesn’t fully explain why global media researchers aren’t making better use of digital media for research purposes. Goggin and McLelland addressed this issue specifically when they argued the internationalization of the Internet offered unprecedented abilities to create a greater global scope for media studies, both in research and in pedagogy (Goggin & McLelland, 2009). They further argue that the multiple types of online content (different formats, channels, applications, and so on) have quite different functions and uses across the world, and it is imperative that media studies move beyond simple studies of single technologies to grasp the way in which media and technology have become intertwined. That is, it is possible to use the Internet to study global media, but it is important to remember the historical, social, and cultural factors which contribute to understanding that media content. We believe that there are at least two key reasons that the rise of digital media has not led to more global media research. First, although there is a great deal of online content available via digital media, global broadcast media are typically accessible only via subscription satellite service, as most US cable systems (as well as a good number of other Western systems) don’t include these channels as a part of their normal offerings. Al Jazeera English, which has attempted to position itself as an important global counterpoint to CNN, remains largely unavailable in the United States as of the time of this writing (July, 2012), with only a handful of local cable companies offering the channel. Although these global broadcast outlets do in fact provide streaming video content online, for many US or UK based media studies scholars, they remain ‘‘under the radar,’’ as it were. Second, and more importantly, even when global media are available, either through satellite television feeds or the World Wide Web, there are a limited number of trained scholars with the language skills to access the content available via these important global media outlets. This has led to an unfortunate and inexcusable gap in our research outputs. Although less true in the case of Spanish language media, this is especially true for media from the so-called ‘‘critical languages’’ (such as Arabic, Chinese, Farsi, Hindi, Urdu, etc.), as the number of scholars engaging these media outlets remains remarkably small.1 Our argument thus far has been that media research remains largely hindered in attempts to internationalize analysis, and there are reasonable explanations for this lack of sustained research on global media outlets. Our purpose in this article, however, is to demonstrate the ways new technological developments are creating potential for new methodologies within global media research, and to illustrate possible ways these technologies can support research, as well as education, on global media. Specifically, we argue that the development of digital satellite capture, natural language processing, and digital translation, present the possibility of innovative methodologies both for research and for teaching in global media.

10



In the following section, we both detail certain key technological approaches to capturing global broadcast content, and introduce the Global Networked Media Archive (GNMA) as an example of one key locus of research capacity, deploying advanced technologies to capture and analyze large amounts of data from a variety of understudied language communities and global media outlets. The GNMA also represents a significant step forward in pedagogical practice, by allowing students, who typically do not have extensive foreign language skills, to access global media in target languages. This system offers unprecedented ability to better understand, analyze, and teach about global media and creates possibilities for new research methodologies to be developed.

The Rise of Foreign Language Media Monitoring Technologies Accessibility to foreign language materials in the absence of foreign language skills is a significant problem not just for academic researchers and students, but also for governments, especially when access to materials is important for national security purposes. After the perceived intelligence failures that led to the September 11, 2001 attack on the United States, the U.S. government began to invest heavily in programs and technologies that would enhance the foreign language capacity of government agencies, as well as personnel. For example, programs to support foreign language study, through the National Security Education Project (NSEP) and the Flagship Language programs, grew in importance after 9/11, in an effort to develop more US citizens with a working knowledge of key critical languages (Kuenzi & Riddle, 2005; National Security Education Program: 20th Anniversary Review, 2011). Likewise, policymakers and others began to research the development of technologies that would also provide immediate access to global media agendas and programs. Interestingly, the narrative of the development of the Internet in some ways links back to the development of these technological innovations that can help to redress the research gap into understanding of global media. The role of DARPA (Defense Advanced Research and Projects Agency) in the development of the Internet is well-chronicled (Naughton, 1999), and the origins of contemporary media analysis technologies have similar roots. DARPA’s mission is to conduct advanced research to meet the needs of the U.S. Department of Defense, and it was this mission that led DARPA to develop a research agenda on machine translation of foreign languages. The overriding need identified by the Department of Defense was for technology that would help to ‘‘close the language gap—to make relevant information, regardless of source language, accessible to Englishspeaking personnel’’ (Olive, Christianson, & McCary, 2011). The agency began to conduct research as far back as 1971 that would contribute to technologies for media monitoring and extraction, but began to accelerate in the later part of the 1980s. The DARPA GALE (Global Autonomous Language Exploitation) project, which began in earnest around the year 2005, made significant breakthroughs



in Natural Language Processing for both Speech to Text Recognition, Machine Translation and Information Retrieval (Olive et al., 2011). In addition, research results and improvements in computer processing made it possible to initiate the transition of research capabilities into operational engines and build systems that support users who need access to the media but do not have targeted language skills. These new engines transitioning research into operational environments initiated the media monitoring systems that make multimedia access and manipulation available to users working on devices as simple as desktop computers (Egan & Sears, 2011). Building upon and extending this work, the Technology Support Working Group (TSWG), a U.S. interagency office in the Department of Defense, sponsored the development of systems to transition the DARPA research into workable prototypes and to overcome the significant gaps in automatic data retrieval, from various media sources, that weren’t dependent upon humans having to watch television all the time (Egan & Sears, 2011). TSWG worked with scientists and engineers at Raytheon BBN to develop a suite of technologies for media monitoring, including technologies for monitoring broadcast networks (including television and radio), Web sites (including blogs, news sites, and governmental sites), and other emerging types of digital media, such as social media. ‘‘Media monitoring’’ in this sense is only loosely related to the term as used in media studies, which focused on appraisals of the ways in which media covered various issues, with the goal of determining the accuracy and fairness of that coverage (see for example, Nordenstreng & Griffin, 1999). Rather, media monitoring within the GALE and TSWG programs focused on developing methodologies to capture large amounts of media content, with the ability to quickly sort through that content to grasp emerging trends, topics, and issues that would have relevance to various governmental agencies. As the technology has been successfully deployed for US governmental clients in the past five years, TSWG has further increased the dual use capacity of the systems and began to develop additional user communities in the academic community, such as in language instruction (Egan & Sears, 2011). What these technologies offer the media research and educational communities is the ability to access important sources of global media, especially for researchers and students without the linguistic skills to access these outlets otherwise. As the system provides a machinegenerated transcript in the original language and from that transcript generates an automatic translation, it allows researchers to watch the original content, in context, to see at the same time the transcript in the original language, and alongside a rough English translation of the content. Figure 1 provides a screenshot of a clip from Al Arabiya, an influential Saudi-based broadcaster. Moreover, because the system creates an automatic storage space for this media, and all the content is automatically ‘‘tagged’’ by the language generated in the transcript, the transcript and translation automatically serve as searchable indexes of a broadcast, regardless of the genre or content of the broadcast. Currently, a system built by Raytheon BBN and deployed at Texas A&M University consists of six separate feeds from three different languages—Arabic, Chinese, and Spanish. The Texas A&M system, dubbed the ‘‘Global Networked Media Archive’’

12

Figure 1 Screen shot of Al Jazeera, November 30, 2012, 3:20 p.m. CST (printed with permission, color figure available online)




(GNMA), creates and stores a 24 hour a day, searchable archive of broadcasts, and is used in media research and analysis, and for coursework related to global media, so undergraduate and graduate students can study global media broadcasts in real time, and is accessible at mms.tamu.edu. Underlying these systems are a variety of distinct technologies, including technologies to harvest foreign language media content, generate machine transcriptions from spoken text (Speech To Text, or STT), assign speaker identity on the basis of voice quality (Speaker Tracking), define the beginnings and endings of text (Sentence Boundary Detection), identify common names and organizations (NamedEntity Recognition), generate machine translations (MT) of text and of speech, distill information (or synthesize information and distill it into coherent and complete English representations that could be used by analysts not conversant in the target language), evaluate and optimize machine translations, and develop operational engines to coordinate the entire process. Given the number of different processes, it is apparent that transcription and translation errors are present in the final output. For example, the system first has to decode the target language broadcast, through the STT process. That output is then subjected to machine translation (MT) and finally, distilled and rendered into standard English. As each of these processes is subject to a certain amount of error, error rates increase towards the tail end of the process. However, the reliability of many of these technologies has improved to an acceptable error rate for confidence in the quality of the output, as measured in reference to Human-targeted Translation Error Rate (HTER) or the degree of completion required by a human translator to capture the full meaning of the text (Liberman, 2008). Accuracy can be measured at numerous points in the process, but it is most important at the STT and MT phases. Current technologies have improved STT word error rate of the GALE system for conversational Arabic to lower than 20%, while for more formal language (such as in broadcast news), it is lower than 15.5%. In Chinese, the character error rate for conversation is at about 20%, while for broadcast news it is 5.4% (Srivastava et al., 2012). The differences between the conversational error rates versus the broadcast news error rates can be understood by thinking of the differences between watching a broadcast drama—which is more likely to use idioms, highly exaggerated inflections, and novel uses of language—and broadcast news, which is more likely to use a standardized delivery and highly regularized language. The second key process is the Machine Translation (MT) phase, and currently, the error rate for Arabic is less than 5% for broadcast news (or highly structured speech) and approximately 10% for unstructured speech, or more conversational speech. Mandarin error rates are higher, at approximately 12% for structured speech, and around 18% for unstructured, conversational speech (Przybocki, Le, Strassel, Moore, & Leusch, 2011; Srivastava et al., 2012). Error rates go down significantly when the text is highly structured, such as in a news broadcast. Overall, researchers with the GALE project found that the system generated HTER rates of less than 10% for broadcast conversation and news in Arabic, while for Chinese the HTER rate is 10% for broadcast news, and 15% for broadcast conversation (Przybocki et al.,


14


2011). Although the system does not have the accuracy of a human translator, these accuracy rates provide reasonably accurate translations for many types of media research, although not accurate enough for understanding nuance, irony, or subtle shades of meaning. Moreover, transcription errors are more likely to occur with newer words or terms, which have not yet been incorporated into the transcription algorithms. When this occurs, the system renders it in the closest terms that it is configured to recognize. Researchers at Texas A&M, for example, have discovered that the now-common term ‘‘Facebook’’ is not one the system recognizes, so when the English term is used on Arabic language broadcasts, it is transcribed as ‘‘Weiss Bok’’ or ‘‘Alves Bok’’ indicating that the system recognized it as a foreign term, and renders it in the closest phonemic element that is recognized. The presence of errors in transcription and translation doesn’t preclude the possibility of developing extensive research programs and agendas using these broadcasts. An untrained observer, with no language capacity in the target language, such as Arabic, can watch a news broadcast from Al Jazeera, understand the topic under discussion, the key organizations or people being discussed, and the conclusions and tone of the discussion, although they would likely be unable to determine subtle variations in language use or argumentation style. There are multiple levels of potential analysis of media content, such as studies on media agenda setting, news framing, visual cues, etc., and many of these do not require perfect translations to conduct. One technique for improving the accuracy of translations is to use the automated systems to find the relevant data (using keyword searches, for example, to identify all the relevant clips on a certain topic), then human translators can improve the translation, thus enabling more confidence in analysis. Thus, the fact that the translations cannot be considered fully accurate should not dissuade innovative uses of this type of system to generate new research agendas and methodologies.

Research and Pedagogical Implications of the GNMA Although the GNMA system archives material for one year, a historical archive of broadcast feeds has been created for content recorded from December 19, 2010 to December 19, 2011. These archives contain the broadcasts from all of the channels capturing several momentous world events, including the breakout of popular protest movements in the Arab world (the Arab Spring), and the March 2011 Sendai earthquake and tsunami in Japan. This special archive is unprecedented in terms of creating an online, accessible archive to historically important media content, such as Al Jazeera’s coverage of the Arab Spring events, data which would be very difficult to find elsewhere. Given the unique access to global broadcast media, these systems provide innovative and unique opportunities for research. The near–real time feed (the system runs about 5 minutes behind real time, in order to provide the necessary transcription and translation) provides an unprecedented access to global media, in both its original



and in translated forms. It allows researchers to look at what is being broadcast at any given moment and allows a keyword searchable archive of past broadcasts, embedded within the actual video stream itself, in a way that rarely exists among broadcast media. Moreover, the system creates a tag-able cross-referencing archive that allows for easier coding and comparison of electronic broadcasts, providing opportunities to create and share native language content about significant global events, or even of just regularly scheduled programming. Finally, the system presents new possibilities for combined textual and visual content analysis of global media broadcasts, allowing researchers to move beyond solely textual analysis of global media transcripts. The ability to extract video clips, along with real time transcripts and translations means that it is possible not only to research global broadcast content but also to present it in a variety of academic forums. The advantages over alternative methods of studying global media are clear. Although outlets such as Al Jazeera and CNN maintain a streaming video feed as well as archives of key stories, the GNMA system keeps the clips within their production and real time broadcast context. Moreover, online archives rarely provide translations of their broadcast, leaving the content largely inaccessible to those without language skills, and without the assistance of an interpreter. There are at least two organizations that provide translated media clips from the Arab world (the Middle East Media Research Institute at memri.org, and Meedan at news.meedan.net), but at least one of these organizations (MEMRI) is often criticized for having overt political goals in the selection of translated materials. But even more important, in the very act of pulling and posting clips, the items have become disembodied from the actual broadcast stream. For example, although Al Jazeera English provides clips of their broadcasts online, Al Jazeera English and Al Jazeera Arabic are functionally separate units, with different production staff, audiences, news agendas, and coverage, even when focused on the same topic. Thus, getting content from Al Jazeera English will not allow the researcher to understand what is happening with Al Jazeera Arabic. Media monitoring technologies, because they combine text, visual, and audio components, also allow for new possibilities for textual and visual content analysis of global media data, as well, which is currently available only through capturing visual content and developing annotated video clips, a time-consuming and labor intensive process. This possibility allows research in international communication to include new elements that are not traditionally studied across cultural boundaries, such as the impact of visual image on audiences and broadcast content. Moreover, the system provides a mechanism for generating comparison between traditional broadcast to new media sources, such as through social media or Web sites, enabling new research agendas in international media studies, such as cross-platform agenda setting and the ways in which broadcast news might be amplified through new media formats. Finally, unlike much social scientific research, which relies on publicly available datasets, international media studies relies heavily upon the interpretations of individual researchers who describe their data and then analyze it, with the scholarly


16


audience having little to no ability to refer back to the original data set. Although researchers do their best to accurately convey content and then analyze it, media monitoring technologies allow the rapid development of online, searchable archives of global media content, tagged and with relevant metadata and annotations, allowing much broader access to a larger community of scholars, and the possibility of creating collaborative data collections, which would enable numerous researchers to access broadcast archives, and develop unique and innovative mechanisms for creating crowd-sourced research projects. Pedagogically, the broadcast monitoring system opens up a new world of possibilities for students. Undergraduate students at Texas A&M have been using this system since 2009, allowing students with no previous exposure to global media to begin to understand the very different news agendas, production values, and expectations of these global broadcasters. Students have deployed the system to do research on topics as varied as Al Jazeera’s coverage of countries in which the Arab Spring was relevant, to CCTV coverage of the Sendai earthquake, comparative analysis of the role of women in various global broadcasters, the death of Muammar Gaddafi, how the various broadcasters covered events such as the volcano in Iceland, the royal wedding of Prince William and Kate Middleton, and a host of geopolitical conflicts. From an educational point of view, this firsthand experience with global media allows students an opportunity to see global news events from a vastly different perspective, and come to very different conclusions about the role and value of global news. Likewise, students have been able to watch a variety of programming, from game shows and comedies on Telesur to costume dramas on CCTV, and develop a critical perspective on these broadcast genres that they would not gain by just watching English-language media. As an example of the type of student learning that is possible with these new technologies, an undergraduate student at Texas A&M conducted a study of Al Jazeera’s coverage of social media in the Arab Spring in relation to the revolutions and protests of the ‘‘Arab Spring’’ within its broadcast media coverage. This content analysis project identified narratives being used by Al Jazeera to frame social media and make claims about the influence they had on the protests and related social movements, and found that Al Jazeera consistently attributed agency to social media, ascribing it a major role in the political movements (Campbell & Hawk, 2012). It is doubtful that students could conduct this kind of research without access to tools that provided this kind of access to long-past broadcast streams.

Conclusion This article seeks to contribute to the agenda of this special issue, which is to develop meta-level thinking on questions of method/methodological approaches presented by both media globalization and the rise of digital media, especially as embodied in media monitoring technologies such as the Global Networked Media Archive. We have argued that in spite of a desire to increase our understanding



of global media, the media studies community is hindered by the lack of access to global media, and by linguistic barriers that prevent engagement with global media. Thus, theorizing about global media without data-driven research will continue to be of limited benefit to the larger community of media scholars. Given the critical importance of global media to larger geopolitical realities, the exclusion from scholarly analysis of global media content is no longer a luxury that the discipline can afford. We have further argued that the rise of new technologies contributes greatly to our ability to develop innovative new methodologies for research in global media. The media monitoring technologies embodied in systems such as the GNMA provide a mechanism by which global media researchers can develop new methodological approaches, by allowing access, retrieval, and analysis of global media to researchers without fluent language skills in a target language. One such methodology might include the ability to ‘‘map’’ a media event according to its global coverage, rather than being limited solely to the media outputs of one language, such as comparative analysis of media coverage of key world events. These approaches would likely find extensive utility in the ability to quickly locate media content across a 24hour broadcast cycle in multiple languages, and thus facilitates the ability to create a cross-referenced archive allowing for easier coding and comparison of international broadcasts. Another methodological innovation would be the ability to track the development of news agendas, media coverage, and production practices across national and linguistic borders, such as comparing the syndication and news value processes through three different national-level broadcasters of global events. This sort of crossnational analysis is currently very limited, but would prove invaluable in developing a sophisticated understanding of the production processes of global news. Likewise, it offers the ability to track news agendas and newsworthiness across national boundaries. Moving beyond just news broadcasts, methodologies that can track the international reproduction of media production practices in genres of entertainment, such as comedies, game shows, or reality shows would be tremendously helpful, and is currently rarely attempted. There is significant work to be done to develop these new methodologies for their greatest effectiveness within the field of international media studies. For example, although access to global media with a translation accuracy rate of 90% is demonstrably better than no access, we need to develop methodological tools that would enhance the confidence in analysis of that content, and to develop levels of analysis that are less prone to error. Likewise, we need further tools that enable the crossnational analysis of visual and audio elements of global media not easily captured in texts. For example, tools that can automatically capture, identify, and tag visual elements (such as a picture of political leader, or a symbol) in the same way that current technologies can capture words, phrases, and other textual elements, would be invaluable in understanding the international and intercultural appropriation of visual elements in global media. Another critical need is for tools that allow for the automatic detection of information as it is transmitted across multiple platforms,


18


such as when broadcast news headlines become tweets, or Facebook updates. This would allow us to see in real time how news is transmitted and transformed through user interaction, and to find ways to conceptualize the ways in which technological innovation and media are intertwined, and allow media studies to move beyond simple studies of single technologies. Regardless of these limitations, it is clear that media monitoring technologies do indeed offer the potential for significant new methodological and pedagogical breakthroughs in global media research. Further development of these technologies promises to expand not only the number of media studies scholars with access to global media, but also the ability of that community of scholars to share their work, and ultimately, contribute to the larger geopolitical conversation about our interactions across national and linguistic boundaries.

Note 1 ‘‘Critical Languages’’ are those deemed by various U.S. governmental agencies to be those for which there is a strong demand for speakers but little supply. Although the list of critical languages varies from agency to agency, it usually includes languages which are deemed critical to the effectiveness of that particular agency, and almost all of the lists include the ‘‘major’’ languages of important geographical regions, such as Arabic, Chinese, Hindi, etc.

References Appadurai, A. (1996). Modernity at large: Cultural dimensions of globalization. Minneapolis, MN: University of Minnesota Press. Banerjee, I. (2009). Asian media studies: The struggle for international legitimacy. In D. K. Thussu (Ed.), Internationalizing media studies (pp. 165–174). New York, NY: Routledge. Best, S. J., Chmielewski, B., & Krueger, B. S. (2005). Selective exposure to online foreign news during the conflict with Iraq. The Harvard International Journal of Press/Politics, 10(4), 52– 70. doi: 10.1177/1081180x05281692 Campbell, H., & Hawk, D. (2012). Al Jazeera’s framing of social media during the Arab spring. CyberOrient, 6(1). The CCTV’s International news channel CCTV-9 is renamed CCTV News. (2010, April 26, 2010). China Screen News. Retrieved from http://china-screen-news.com/2010/04/thecctvs-international-news-channel-cctv-9-is-renamed-cctv-news/ Chakrabarty, D. (2007). Provincializing Europe: Postcolonial thought and historical difference. Princeton, NJ: Princeton University Press. CNN. (2012). CNN.com-Contact Us. Retrieved November 28, 2012, from http://www.cnn. com/feedback/cnni/ Curran, J., & Park, M.-J. (Eds.). (2007). De-westernizing media studies. London: Routledge. Downing, J. D. H. (2009). International Media Studies in the U. S. Academy. In D. K. Thussu (Ed.), Internationalizing media studies (pp. 267–276). Abingdon, UK: Routledge. Egan, K., & Sears, A. (2011). Bridging the gap between research advances and operational users. In J. Olive, C. Christianson & J. McCary (Eds.), Handbook of natural language processing and machine translation (pp. 846–860). New York, NY: Springer. Ekecrantz, J. (2009). Media studies going global. In D. K. Thussu (Ed.), Internationalizing media studies (pp. 75–89). New York, NY: Routledge.


Kluver, Campbell, and Balfour/LANGUAGE AND THE BOUNDARIES OF RESEARCH 19 Goggin, G., & McLelland, M. (2009). The internationalization of the Internet and its implications for media studies. In D. K. Thussu (Ed.), Internationalizing Media Studies (pp. 294– 307). New York, NY: Routledge. Halavais, A. (2000). National borders on the World Wide Web. New Media & Society, 2(1), 7–28. doi: 10.1177/14614440022225689 Kluver, A. R. (2002). The Logic of new media in international affairs. New Media and Society, 4(4), 499–517. doi: 10.1177/146144402321466787 Kuenzi, J. J., & Riddle, W. C. (2005). National Security Education Program: Background and Issues. Washington, DC: Congressional Research Service. Retrieved from http://www.fas.org/ sgp/crs/misc/RL31643.pdf Liberman, M. (2008). HTER. Retrieved from http://languagelog.ldc.upenn.edu/nll/?pD193 Lynch, M. (2006). Voices of the new Arab public: Iraq, Al-Jazeera, and Middle East politics today. New York, NY: Columbia University Press. MacBride, S. (1980). Many voices, one world. London and New York: International Commission for the Study of Communication Problems. MacKinnon, R. (2004). Priorities of American global TV: Humanity, national interest, or commercial profit? Paper presented at the 14th JAMCO Online International Symposium. National Security Education Program: 20th Anniversary Review. (2011). Washington, DC: Retrieved from http://www.nsep.gov/about/support/20%20Year%20Anniversary%20Review. pdf Naughton, J. (1999). A brief history of the future: origins of the Internet. Great Britain: Weidenfeld & Nicholson. Nordenstreng, K., & Griffin, M. (Eds.). (1999). International media monitoring. Cresskill, NJ: Hampton Press, Inc. Olive, J., Christianson, C., & McCary, J. (Eds.). (2011). Handbook of natural language processing and machine translation. New York, NY: Springer. Przybocki, M., Le, A., Strassel, S., Moore, C. T., & Leusch, G. (2011). GALE machine translation metrology: Machine translation assessment. In J. Olive, C. Christianson & J. McCary (Eds.), Handbook of natural language proessing and machine translation (pp. 795–811). New York, NY: Springer. Saad, L. (2002). Al-Jazeera viewers perceive west differently. Retrieved from http://www.gallup. com/poll/5860/aljazeera-viewers-perceive-west-differently.aspx Sabry, T. (2009). Media and cultural studies in the Arab World: Making bridges to local discourse of modernity. In D. K. Thussu (Ed.), Internationalizing Media Studies (pp. 196– 213). New York, NY: Routledge. Seib, P. (2008). The Al Jazeera Effect: How the new global media are reshaping world politics. Washington, DC: Potomac Books. Srivastava, A., Colbath, S., Saikumar, G., Watson, J., Lillie, M., & Natarajan, P. (2012). Technology translation in GALE. Cambridge, MA: BBN Technologies. Tehranian, M. (1999). Global communication and world politics: Domination, development, and discourse. Boulder, CO: Lynne Rienner. The 2011 Arab Public Opinion Poll. (2011). Brookings Institution. Thussu, D. K. (Ed.). (2009). Internationalizing media studies. London and New York: Routledge. Volkmer, I. (1999). News in the global sphere: A study of CNN and its impact on global communication. Luton, Bedfordshire, UK: University of Luton Press. Yuan, E. J., & Webster, J. G. (2006). Channel repertoires: Using Peoplemeter data in Beijing. Journal of Broadcasting & Electronic Media, 50(3), 524–536. doi: 10.1207/s15506878 jobem5003_10 Zayani, M. (Ed.). (2005). The Al Jazeera phenomenon. Boulder, CO: Paradigm Publishers. Zhao, Y. (2009). Rethinking Chinese media studies: History, political economy, and culture. In D. K. Thussu (Ed.), Internationalizing Media Studies (pp. 175–195). New York, NY: Routledge.