CGNET SWARA WP

65

Preeti [email protected] CandidateDepartment ofCommunication StudiesUniversity of Texas, AustinWhitis Avenue, 2504AAustin, TX 78712USA

Jonathan [email protected] Research, India9 Lavelle Rd.Bangalore, 560001India

William [email protected] Research, India9 Lavelle Rd.Bangalore, 560001India

Emergent Practices Around CGNet Swara MUDLIAR, DONNER, THIES

Research Article

Emergent Practices Around CGNetSwara: A Voice Forum for CitizenJournalism in Rural India

Preeti MudliarJonathan DonnerWilliam Thies

Abstract

Rural communities in India are often underserved by the mainstream media.While there is a public discourse surrounding the issues they face, this dia-logue typically takes place on television, in newspaper editorials, and on theInternet. Unfortunately, participation in such forums is limited to the most priv-ileged members of society, excluding those individuals who have the largeststake in the conversation. This article examines an effort to foster a more in-clusive dialogue by means of a simple technology: an interactive voice forum.Called CGNet Swara, the system enables callers to record messages of local in-terest, as well as to listen to messages that others have recorded. Messagesare also posted on the Internet as a supplement to an existing discussion fo-rum. In almost three years of deployment in India, CGNet Swara has loggedmore than 137,000 phone calls and released 2,100 messages. To understandthe emergent practices surrounding this system, we conducted interviews with42 diverse stakeholders, including callers, bureaucrats, and media members.Our analysis contributes to the understanding of voice-based media as a vehi-cle of social inclusion for remote and underprivileged populations.

IntroductionThe rapid penetration of mobile phones in the world’s low-income regionshas triggered widespread interest in building mobile systems and applica-tions to beneªt education, health, government, and other social ends.Recently, some projects from the ICTD conference series have emphasizedthe role of users as active participants and producers of information, anal-ogous to Web 2.0 (Agarwal, Kumar, Nanavati, & Rajput, 2009; Banks &Hersman, 2009; Patel, Chittamuru, Jain, Dave, & Parikh, 2010; Sterling,2009). Participatory processes are of particular interest to developmentpractitioners, since they encourage accountability, local ownership, andproblem solving, and these processes situate a broader share of power,decision making, and inºuence with communities, rather than outsiders(Midgley, Hall, Hardiman, & Narine, 1986). And as a medium for increas-ing or enabling participation, mobiles hold particular promise (Goggin &Clark, 2009), due partly to their support of voice communication. Thespoken format transcends literacy, machine-written text/font issues, andlocal language issues, and it is accessible from any handset.

© 2013 USC Annenberg School for Communication & Journalism. Published under Creative Commons Attribution-Non Commercial-Share Alike 3.0 Unported

license. All rights not granted thereunder to the public are reserved to the publisher and may not be exercised without its express written permission.

Volume 9, Number 2, ICTD2012 Special Issue, 65–79

In this article, we offer a casestudy of a citizen journalism net-work, previously reliant on com-puters and the Internet, and weexamine its evolution as it hasadopted a voice-based interfacefor recording and listening toinformation over the phone. Thefocus of our study is CGNet,which targets the Indian state ofChhattisgarh. In the words of itsfounder, “CGNet is the people’swebsite of Chhattisgarh, whereeverybody is a journalist. It is acitizens’ journalism forum whosemission is the democratization ofjournalism, where journalism isnot restricted only to journalists”(Choudhary, 2009). Since 2004,the CGNet website and mailing list have provided aforum for discussion of the Chhattisgarh region,speciªcally for issues related to its development andpeople. While the site and mailing list are active,with more than 230 messages per month, depend-ence on the Internet made it difªcult for ruralcommunities—those most impacted by the issues athand—to access or contribute to the dialogue.Internet penetration in Chhattisgarh stands at 0.5%(India Ministry of Communications & InformationTechnology, 2010).

Out of this ecosystem was born CGNet Swara,1 aportal to CGNet that uses mobile phones to extendparticipation beyond the Internet’s reach. Before weproceed, a note is in order regarding our approachto this article. Our primary tone is one of externalexamination and critique, as the ªrst two authorsplayed no role in the creation of either CGNet orCGNet Swara. However, the third author is a mem-ber of the CGNet Swara technical team. He facili-tated this study by providing call logs, contactdetails, and system analytics to the other authors.While this study was done with the knowledge andconsent of the full CGNet Swara team, the conclu-sions reached do not necessarily represent theirviews.

CGNet Swara users place an ordinary phone callto the system, which presents them with twooptions: Press “1” to record a message, or press

“2” to listen to other messages (see Figure 1). Atrained journalist moderates the recordings prior topublishing them on the channel. Published record-ings are accessible via both the phone and awebsite,2 where they are also summarized in textform. The CGNet Swara staff often seeks to spuraction on the reports by disseminating them to con-tacts in the mainstream media, as well as in the gov-ernment. In the ªrst 34 months of its deployment,CGNet Swara featured 2,100 reports and loggedmore than 137,000 calls into the system. As detailedin the coming sections, some reports have led to theredress of important grievances that beneªted sev-eral communities.

Our goal in this article is to understand theusage, perceptions, and impact of this platform dur-ing the period when it transitioned from CGNet toCGNet Swara. We start by characterizing the sys-tem’s content through conducting an analysis ofmore than 1,000 voice recordings. Then, we exam-ine the social ecosystem that created (and was cre-ated by) the platform via interviews with 42 diversestakeholders, including contributors, listeners, andthe moderator, as well as such external actors asjournalists and government authorities. Based onour analysis of these conversations, we identiªedthree themes: 1) the experience of users interactingwith the voice interface, 2) the emergent practice ofusing the platform as a mechanism for grievance

66 Information Technologies & International Development

EMERGENT PRACTICES AROUND CGNET SWARA

1. Swara is the Sanskrit word for voice.2. http://cgnetswara.org/

Figure 1. CGNet Swara ecosystem.

reporting, and 3) the complex relationship betweenCGNet Swara and established media outlets. Weconclude by positioning our inquiry into CGNetSwara at the intersection of overlapping but distinctacademic discourses in addition to ICT4D, spanningcitizen journalism, community informatics, ande-governance/transparency.

Related WorkThe devices we call mobile phones offer varied bun-dles of features and affordances, made more power-ful by both their networks and the off-deviceprotocols running on servers in data centers. Tech-nologists and development practitioners can (andhave) built applications to take advantage of SMS,USSD, MMS, still photos, video upload and play-back, and the broader Internet via the data channel(Donner, Verclas, & Toyama, 2008). Their decisionsinvolve oft-debated tradeoffs among affordability,richness, and reach. And yet every handset on theplanet has a microphone and a speaker, and voice isas close to a universally shared approach to commu-nication as humans have. Voice is not a panacea; itis linear and hard to search, languages are frag-mented, and phone calls can be costly relative toother messaging channels. But in many circum-stances, voice interfaces offer ubiquity, affordability,and ease of use (Goggin & Clark, 2009; Patel,2010).

Thus, it is encouraging to see a burst of voiceapplications tackling a variety of development prob-lems in a variety of resource-constrained settings.Examples include giving voicemail to homeless indi-viduals in the United States (Le Dantec et al., 2011),automated answering machines for communityradios in India to build engagement (Koradia & Seth,2012), offering ad hoc group support for NGOs(Odero et al., 2010), delivering health information(Sherwani et al., 2007), bringing Web content toresource-constrained classroom settings (Ford &Botha, 2007), and exchanging agricultural advice for(Plauché & Nallasamy, 2007) and by (Patel et al.,2010) farmers in India.

Some voice projects stress navigation of voicemenus, others leverage speech recognition, and stillothers try to capture and retain contributors’ voices.These last systems are perhaps the most interestingfor those seeking to facilitate participatory processesin communities where low literacy or other factors

reduce access to or use of other media (from lettersto the editor to blogging). Speciªc voice-basedimplementations have begun to blur the linesbetween mobile handsets and community radio,allowing communities with little or no capacity tocreate text-based content to be both listeners to andproducers of the “stories” they want to tell and theagendas they want to set (Bailur, 2007; Frohlich etal., 2009; Kotkar, Thies, & Amarasinghe, 2008;Ranganathan & Sarin, 2011; Slater, Tacchi, & Lewis,2002; Sterling, 2009).

Still, many of these community platforms remainpilots or experiments. Deployment of broader sys-tems for sustained use in the ªeld remains rare. As a2010 review of citizen media by the Open SocietyFoundation suggested:

Very few of the projects we documented reportback to citizens via text or voice systems aboutthe information they have collected. Aside fromCGNet Swara, only the Budget Tracking Tool inKenya and the African Elections Project allow citi-zens to access, rather than only report, informa-tion via non-Internet-connected phones. (Avila,Feigenblatt, Heacock, & Heller, 2010, p. 35)

Evolution of CGNet SwaraThe geopolitical context for this work is Chhattis-garh, a small state in central India that was formedin 2000. Chhattisgarh is populated primarily by theAdivasis, an indigenous people who are among thepoorest and most socioeconomically disadvantagedin all of India. Of the state’s 25 million inhabitants,80% live in rural areas, and 30% are illiterate. Thearea is also home to the Maoist insurgency, a violentleft-wing movement. In 2007, Prime MinisterManmohan Singh designated this insurgency asIndia’s greatest internal security threat.

Chhattisgarh is a difªcult environment for themedia. Due to a shortage of trained journalists inrural areas, there are no established news sources inthe local tribal languages, such as Kurukh or Gondi,each of which has more than 2 million speakers.The situation is worsened by regulations surround-ing community radio in India. In addition to the highcost of establishing a community radio station, it isillegal to broadcast or discuss news via communityradio. While newspapers and television stations havea presence in the state, only rarely do they cater tothe needs or broadcast the voices of the Adivasipopulation.

Volume 9, Number 2, ICTD2012 Special Issue 67

MUDLIAR, DONNER, THIES

CGNet addresses this need by providing a plat-form for tribal community members to report anddiscuss issues that are meaningful to them. Reportsfrom such citizen journalists ºow through personalcommunication to a CGNet moderator, who broad-casts their stories for discussion on the CGNetwebsite and mailing list. Since its establishment in2004, CGNet has been recognized (ManthanAward, 2008) as fostering dialogue and effectingchange within the state. As of July 2011, CGNet’smailing list had more than 2,000 members. How-ever, due to dependence on computers and theInternet, most of those members were drawn fromthe urban, English-speaking elite.

CGNet Swara’s goal is to extend CGNet’s reachto anyone with access to a low-end mobile phone.As described previously, callers record stories and lis-ten to other recordings by navigating a simple inter-active voice response (IVR) system. Recordings,which can be a maximum of three minutes, undergomoderation to ensure they are clear, audible, andappropriate for dissemination. Once the moderatorapproves a post, it is available for listening on boththe phone and the Internet website. The websitealso includes the moderator’s textual summary ofeach post (typically translated to English, thoughoccasionally left in Hindi) to facilitate search andbrowsing. To keep the phone line available for multi-ple callers, only the four most recent posts are avail-able for playback on the phone. These posts areplayed in time order, with the most recent ªrst; call-ers may skip to the next post by pressing a key. Cur-rently, there is no ability to search or browse olderposts via the phone, though the textual summariesof posts are searchable on the website.

The technology underlying CGNet Swara is sim-ple. Following a vision established years ago (Kotkaret al., 2008), a Linux server uses Asterisk (an opensource telephony platform) in combination withLoudBlog (an open source audio blogging platform)to provide the key functionality. The IVR system logicis written in Python and is available as a free, opensource download. The server uses three GSMmodems (Topex Mobilink IP), available for aboutUS$500 each in India, to interface with mobile SIMcards. The server also connects to a pair of BRI digi-tal landlines (supporting two parallel calls each) to

support the original phone number publicized forthe system.

Airtime cost is an important consideration for anysystem using voice calls to report and disseminateinformation. During the ªrst six months of theCGNet Swara deployment, the phone call costswere supported by the callers themselves. However,as the server was based in Bangalore, this consti-tuted a long-distance call from Chhattisgarh, andCGNet staff were concerned that this (temporary)geographic anomaly was artiªcially stunting serviceexpansion. Thus, starting in August 2010, calls toCGNet Swara became free for a limited period (as ofthis writing, it is still ongoing). Rather than using atoll-free number, which can be costly in India, callersare asked to send a “missed call” to the server. Tosend a missed call, users dial the Swara number andhang up while the phone is still ringing. The serverthen returns the call immediately. Since answeringthe call is free and sending missed calls is common-place in India (Donner, 2008), this incurs no inconve-nience or expense to callers. The cost borne by theserver is approximately Rs.0.60 (US$0.012) perminute.

Deployment and Usage StatisticsCGNet Swara was deployed in February 2010. Pre-ceding deployment, 29 students, social activists, andChhattisgarh residents attended a two-day trainingcamp. CGNet’s founder, a former BBC journalistwho also serves as the principal moderator forCGNet and CGNet Swara, led the instruction. Thecamp included an introduction to citizen journalism,as well as practice sessions on interacting with thetechnology. While this event was important to raisevisibility of the service, of the 20 trainees whoowned cell phones at the time, only two were activeCGNet Swara users six months later. Thus, aware-ness and participation in the system spread mainlyby word of mouth.

As of December 2012, CGNet Swara hadreceived 137,000 phone calls and posted 2,100recordings (see Figure 2). Currently, it publishesabout three new posts and receives approximately200 calls per day; thus, the vast majority of callersonly listen to content. System posts were contrib-uted by at least 715 distinct callers.3 As in many sys-tems hosting user-generated content (Ocha & Duval,



3. Caller ID was unavailable for the project’s ªrst six months. Thus, the number of distinct callers is actually higher thanreported.

2008), the 10% most active contributors are respon-sible for a large fraction (51%) of the posts. Morethan 18,000 distinct callers have listened to content;of these, 2,850 are “regular” callers (they havecalled 10 or more times), while 7,500 people calledthe system only once. The 10% most active listenersare responsible for 61% of the phone calls. Theaverage phone call is three minutes long, and theserver streams approximately 10 hours of audio con-tent per day. The system is growing steadily, withroughly 25 new contributors authoring a post eachmonth, and about 700 new listeners calling in forthe ªrst time.

Content of PostsTo characterize the content on CGNet Swara, weinformally categorized all the stories published dur-ing the ªrst 20 months (n � 1,012). Figure 3 showsthe results by content type, while Figure 4 showsthe breakdown by subject matter. Apart from news,grievances were the most common type of post,mostly related to a variety of livelihood- and civic-related issues. They constituted 34% of all posts andranged from complaints over disparity in insurancerates received by farmers to a demand for betterwages for laborers. Nonpayment of wages underthe National Rural Employment Guarantee Act(NREGA) formed the bulk of grievances. Enacted in2005, the law guarantees 100 days of manualunskilled labor per year, paid at a rate of Rs.100(US$2.00) per day, to willing adults.

Some types of recordings wereunexpected. For example, reportscategorized as “performances”represented song and poetry,often rich in cultural tradition.Consistent with unsolicited per-formances observed on a differ-ent voice forum (Patel et al.,2010), it may be that the univer-sal unmet need in rural India isthe desire to sing! The moderatordid not discourage such submis-sions, as celebration and preser-vation of tribal culture is fullywithin the scope of CGNetSwara’s goals. Often, these mes-sages also had social overtones.

To gain additional insight, weperformed a more detailed coding for two monthsof reports (December 2010–January 2011; n �

110). During this focus period, 85% of posts werein Hindi, 10% were in Kurukh, and the rest weremade in other tribal languages. To the best of ourknowledge, this represents the ªrst source of aggre-gated news in the Kurukh language. Looking for-ward, the CGNet Swara team aspires to solicit morecontent in tribal languages.

During the focus period, 69% of contributorswere male. Most (90%) of the contributors revealedtheir name, while the rest claimed to be speaking onbehalf of an organization. Around 82% chose toreveal their location. Almost 70% of the content con-centrated on localized issues, as compared to issuesof signiªcance at a state or national level. Almost20% of posts pertained to NREGA grievances. A simi-lar number (21%) of the contributors were interven-ing on behalf of someone else’s grievances, andaround 45% of the contributors were reporting per-sonal experiences or eye-witnessed events.

Anecdotal ImpactThere have been several cases in which reports onCGNet Swara have led to a measurable impact onlocal communities. Usually, this impact originatedwith actions taken by the CGNet Swara moderatoras he lobbied mainstream media and governmentofªcials to follow up on a given report. Impact sto-ries are detailed on the CGNet Swara website;4

we offer just one example here.



4. The CGNet Swara website (cgnetswara.org/) is distinct from the CGNet website.

Figure 2. Monthly trafªc (calls and posts) to CGNet Swara (the dip in early2010 was due to an outage while the server moved to a new location).

This impact story relates to overdue NREGAwages, a common theme described previously. ACGNet Swara contributor associated with an NGO inSarguja, a northern district of Chhattisgarh, usedthe service to help a man named Pitbasu Bhoireceive the wages due to him. Pitbasu’s situationwas urgent: He desperately needed money to pro-vide medical care for his son. Pitbasu’s case was pre-sented in a series of calls to CGNet Swara. The ªrstcall highlighted his struggle to receive payment; thenext recording informed listeners of Pitbasu’s son’sdeath as he awaited payment for his labor.

The incident, which was brought to light viareports on CGNet Swara, led immediately to a seriesof reports in two of the most widely circulatednewspapers in India (The Times of India and

The Hindu), as well as the BBC’sHindi edition. This publicity, inturn, prompted governmentofªcials to expedite the paymentof Pitbasu’s wages, which weredelivered within two weeks of theªrst coverage on CGNet Swara. Asubsequent recording on CGNetSwara thanked the system forhelp securing Pitbasu’s wages andurged listeners to record morecases of nonpayment, so thattheir grievances could receive sim-ilar redress.

While a comprehensive impactassessment is beyond the scopeof this article, anecdotes such asthis one provide a powerfulexample of CGNet Swara’s poten-tial role in garnering attentionand effecting change with respectto speciªc problems. We exploreelements of this potential, emer-gent role in the rest of this article.

Qualitative Methods

Data CollectionThe data reported in this studywere primarily collected usingface-to-face and telephone inter-views, as well as ªeld observa-tions. One of our goals for thisqualitative research was to under-

stand how users and other stakeholders perceivedand used the system, perhaps in ways that differedfrom the founders’ expectations. Hence, we werecareful to ensure that the questions posed torespondents avoided terms such as news, journal-ism, and information. Instead, respondents wereasked questions such as, “How would you describeCGNet Swara to someone?” and “Who do youthink listens to you on CGNet Swara? Why?”

Study respondents were recruited through thecall logs captured by the CGNet Swara system. Oncecontact was established with an interviewee, weasked for permission to conduct and record an inter-view. Participants were free to opt out of either therecording or the interview at any time. Respondentswere assured of the privacy of the conversation, and



Figure 3. Twenty months of CGNet Swara posts by type. A post may be as-signed to more than one type.

Figure 4. Twenty months of CGNet Swara posts by subject. A post may beassigned to more than one subject.

they were explicitly told that the conversation wouldnot be played back publicly or broadcast on CGNetSwara.

To complement the interviews, we undertook aªeld visit for two weeks to four districts in Chhattis-garh. Contributors were observed and interviewedin a variety of contexts, such as public meetingsorganized to protest against wage nonpayment,sites that were experiencing land acquisition threats,and beneªciaries whose grievances were resolved asa consequence of recording on CGNet Swara. Thebeneªciaries were mostly NREGA workers whosecomplaints about wage delays were resolved afterreporting on Swara and were interviewed duringªeld visits. Taking note of the high number of wage-related (NREGA) grievances, a visit to the complaintofªce set up by the state government was alsoundertaken to interact with the government staffworking there. In addition, we interviewed main-stream news journalists reporting from the statecapital of Raipur, as well as senior bureaucrats of thestate.

We conducted 42 interviews, including one withthe CGNet Swara founder (who also serves as itsmoderator). Of these, 17 were content contributors,14 were listeners, and two were beneªciaries of theservice. We also interviewed three bureaucrats andªve journalists. Seventeen interviews were con-ducted via telephone; the other 25 interviews wereconducted face-to-face. We conducted 34 inter-views in Hindi, while the remaining eight were con-ducted in English.

Data AnalysisAll interviews were transcribed for review and analy-sis; those interviews conducted in Hindi were trans-lated into English. We performed the analysis forthis article by coding the interviews into categoriesto discern distinctive patterns that informed boththe use of the system and the response to it.Through this process of induction, ªve initial broadcategories emerged during mid-ªeldwork debriefsand during transcription and coding: 1) broad usagetrends of the portal, 2) the signiªcance of usingmobile phones to report information, 3) the interac-tion with and usage of an interactive voice portal,4) the emergence of grievance reporting as the mostimpactful use of CGNet Swara, and 5) tensions thatsurfaced as CGNet Swara established itself as analternative communication medium.

In the latter stages of our analysis and discussion,we aggregated these ªve domains into three, whichare presented below. We describe, in turn, users’experiences of interacting with the voice interface,the emergent practice of using the platform as amechanism for grievance reporting, and the com-plex relationships between CGNet Swara and estab-lished media outlets.

Qualitative Analysis

Different ContributorsThe system’s builders sought to allow anyone inChhattisgarh to participate in the practice of jour-nalism. Of 31 callers, 22 provided demographicinformation; the sensitive nature of the contextmost likely led the others to be careful with theiridentities, especially during phone interviews. Of the22 interviewed callers who provided demographicinformation, all were literate and possessed at leastan eighth-grade education. Most (82%) were male.Several respondents worked as social activists,although others worked as laborers, constructionworkers, journalists, or in other occupations. All ofthe respondents read the newspaper, while 72%watched TV news. Internet use was scarce; only twointerviewees mentioned using the Internet as a newssource. Our interviews were neither numerousenough nor randomly selected enough to representall CGNet users. We would like to note that theusers that agreed to be interviewed for this studywere guaranteed anonymity and were assured thatno identifying information about them would beused. From the data that we analyzed, initial indica-tions were that a savvy subset of residents tookadvantage of the system. Hindi-speaking activistsand NGO workers (though generally non-Internet-using ones), rather than disadvantaged members ofthe community, interacted most heavily with the sys-tem. These individuals also enabled intermediatedusage (Sambasivan, Cutrell, Toyama, & Nardi, 2010)by a broader set of people.

The lack of direct, broad-based usage could bejust a matter of awareness. An NGO worker offeredsome suggestions:

Even if I tell 50–100 people about this service in aday, it won’t be an effective way to spread themessage. This can’t be done in a day. You have togive people some time to learn what Swara is.Workshops are a good way to solve this problem.



The journalists were skeptical, suggesting that theproblem with engagement went beyond awareness.Even though none of our study respondentsattested to having felt any kind of persecutory pres-sure for having used CGNet Swara, one journalistreferred to the “shy nature” of the poor person:“Villages in Chhattisgarh are the kind of placeswhere everybody knows everybody else. You couldeasily be tracked and found out and there could berepercussions from having complained about some-one or something.” A senior editor of a regionaldaily added the following:

We cannot expect a weak defenseless person tohave the courage to speak out on Swara. The ser-vice is a good beginning, and it is my hunch thatNGOs are the ones who are making the most useof it. Let it run for a while because it is still in itsinfancy, but it is true that if you are thinkingabout the tribals whom this should beneªt, thenyou have to remember that that individual ªrstlydoes not have the conªdence to speak out. Sec-ondly, talking on the mobile phone to someone orsomething he does not know is a big thing.

Different Affordances, New ContextsThe voice interface is convenient and understood.One respondent put it bluntly, “We just dial andtalk.” Another regular said, “I can present issuesvery easily on Swara and can disseminate it easily,too. I like the fact that I can present issues wheneverI feel at my convenience.” A third emphasized thephone’s portability:

You can listen or record on Swara from anywhere.You don’t have to make time for it. I listen to itevery morning when I walk to the lake to bathe.Once there, I sometimes also share stories that Iªnd interesting with others on the lake by switch-ing on the speaker phone.

The mobile phone’s portability led to the creation ofnovel spaces for interaction around CGNet Swara.Participants were able to act as broadcasters/infomediaries or journalists, generating or sharingcontent with others present. One regular contributorto the service dialed in from a moving train torecord an interview with a group of fellow passen-gers when she learned about their wage paymentgrievances. Another contributor said that he ªrstheard about CGNet Swara when his curiosity wasaroused by seeing people listening to stories in the

local train compartment during his daily commute.Others described switching on the phone’s externalspeaker to enable small groups to listen to stories atthe same time. A contributor who once sent in arecording from a public meeting in a village told hisstory:

Most of the complaints were with regard to civicutilities so I just dialed the Swara number, put iton speaker phone, and made them listen to a fewstories as a demonstration. We then recordedtheir own complaints on Swara.

In this framing, we emphasized the capacity ofbeing able to speak and hear, rather than the sym-bolic or non-instrumental feelings associated with“being heard.” We heard some comments to thiseffect: One respondent reported “feeling very niceafter I heard myself on the Swara service.” Otherssaid they would take pride in being selected for thesite, rallying friends and family to listen, but theseresponses were not as widespread as those aboutthe simple practicality of the interface as a tool foron-the-ºy citizen journalism.

Calling for Grievance RedressA common (but by no means exclusive) use ofCGNet Swara was the pursuit of grievance redress,an individual’s complaint that was identiªed, medi-ated, described, or ampliªed by an activist/journalistagainst an element of the local bureaucracy notdoing what it should. In this way, Pitbasu’s case isarchetypal. Explaining the difªculties that he facedwith government ofªcials before resorting to CGNetSwara, the contributor who intervened on his behalftold this story:

We were trying to get in touch with governmentofªcials and I sent a couple of my colleagues tomeet the concerned ofªcer. He did not listen. Weapproached them a second time, they did not lis-ten even then. We met the commissioner andeven he directed the ofªcer to pay the wages, butthe payment still did not happen. I felt that someaction was needed. I was in Raipur at that timeand a lot of pressure was needed. So I decided topick up the phone and call in a last-ditch attemptto see what would happen.

Contributors’ responses reveal that CGNet Swarais perceived as a tool that carries complaints forwardand helps in their resolution. When we asked one



contributor how she would explain CGNet Swara tosomeone who knew nothing about it, her responsecut right to this role: “I tell them it is a very nicemedium where we can speak about any irregulari-ties in government schemes. When you speak, thenthe government and the administration listen to youand they take steps to address the problem.”

We heard similar framings offered by ofªcials onthe receiving end of the grievances. Even whileexpressing reservations about CGNet Swara beingrepresentative of “the people” (rather than activ-ists), one administrator conceded the utility ofsystems such as CGNet Swara to ensure thatentitlements due to people are disbursed. He sug-gested that such a service could be most usefulwhere corruption was rife, because it had thepotential to usher in more transparency. Anotheradministrator, a former NREGA ofªcial, suggestedthat, were he still in the same post, he would con-sider CGNet Swara to be immensely helpful in track-ing NREGA grievances.

Even more telling, perhaps, is the actual use ofthe CGNet Swara archives (via the website) by atleast one NREGA staffer to track down grievances.This same staffer would prefer that people call theNREGA state helpline number directly (to reduce“bogus complaints”), but nevertheless, the stafferacknowledged the utility of the CGNet Swararesource.

Incidentally, the state’s “ofªcial” NREGA helplinenumber is little-known among Swara contributors,let alone the laborers they often represent. In onlyone of four districts were contributors aware of thehotline’s existence. A caller to CGNet Swara alsorecorded a message asking people to use the help-line number in a bid to boost awareness about it.

Our analysis suggests a particular, almost symbi-otic relationship among the three actors: the com-plainant, the citizen journalist, and the ofªcial.When ofªcials use this system to ªnd and rectifygrievances, everyone beneªts. An editor of a dailynewspaper framed this relationship:

You will see that in most cases the governmentacts like a giver, like a benefactor. So it is all aboutdoling things out. That is the attitude of the gov-ernment, and Swara can succeed here, but as acitizen when you start talking about your rights,that is where the government comes across asweak. Resolving grievances is a different issuethan talking about rights.

Another high-level ofªcial echoed this view:

Suppose I come to know [through CGNet Swara]that this village has not received theirentitlements, then I can intervene or the govern-ment can intervene and see that entitlements arebeing done. Suppose tomorrow they say that myvillage is very nice and I need an airport in this vil-lage then nothing will happen.

The resolution of discrete, actionable grievancesencourages citizens, sustains journalists, and but-tresses the credentials and structural power of theofªcials—although perhaps all this occurs whilebroader demands for more systematic and wide-spread reform remain unsatisªed.

Audiences and Actors in the EcosystemWe asked callers and listeners, “Who do you think islistening?” Some assumed a direct process, with de-cision makers (such as the NREGA ofªcial) listeningdirectly; others perceived links between the voicesystem and the Internet; still others emphasized theinteractions between CGNet Swara contributors andtraditional media. Indeed, each held a part of thepuzzle.

One caller was ªrm in his belief that the chiefminister of the state was a regular listener, “becausehe looks at everything. So obviously, he will be lis-tening to Swara, too.” Another contributor pro-claimed a similar conªdence in CGNet Swara’saudience: “Everybody. Farmers, ‘the big people,’ thegovernment administrators. Everybody listens to it.”Another contributor added the following:

With CGNet Swara, whatever we speak directlygoes on to the net, and it reaches the concernedofªcials. We don’t need to write any applicationor spend any money. We just have to give amissed call and we can communicate throughthat.

But others perceived that CGNet Swara had arole to play in translating or bringing “voices” outto the wider Web. To some, CGNet Swara is simplya means of sending their stories to the outsideworld. In the absence of computer and Internetinfrastructure in the geographies these contributorsinhabit, or of the skills required to access them,CGNet Swara is seen as way to overcome theseobstacles to accessing the Internet. One contributoraddressed the cross-pollination concept directly:



What happens these days is that the stories arealso put on Facebook, and a lot of people arereading it. Some people are taking legal actionbased on the stories, some people are calling upgovernment ofªcers after learning of incidents. Sothis is creating a pressure. CGNet Swara is spur-ring action on stories.

In other cases, contributors perceived that Swarawas not simply a technical bridge to the Internet,but a conceptual and structural bridge to the peoplewho used the Internet. This is due in no small partto the track record of CGNet, the discussion group.Most of the government administrators who wereinterviewed for this study spoke about knowing theCGNet Swara founder through his work as a jour-nalist or as a founder of the Internet forum.

Indeed, most external stakeholders of CGNetSwara, such as journalists and bureaucrats, point tothe website, the service’s emails, its Facebook page,or the Twitter account as more convenient ways toreceive content from the service than calling theCGNet Swara number. The occasional NREGAofªcial aside, most decision makers were more likelyexposed to translated, textual versions of the“voices” of individuals, which had already beenmediated by a citizen journalist and edited/tran-scribed by Swara staff. Says a mainstream journalist,“Checking CGNet Swara is not a regular part of myroutine. I feel that if there is anything important, itwill be emailed to me by the founders anyway.”

But sometimes, CGNet Swara messages areimportant enough to merit attention from the main-stream media, and this attention represents anotherfactor in the multidimensional interactions amongindividuals, contributors, moderators (Swara staff),and decision makers. The contributor who helpedPitbasu get his past-due wages using CGNet Swarasays that the resulting media coverage of Pitbasu’scase caused a huge boost in his morale, leading himto become a regular contributor to CGNet Swara onNREGA issues. Another contributor says that herbelief in CGNet Swara as an effective medium forredress stems from the fact that she was paid a visitby a journalist from a leading English daily to reporton the village’s NREGA status.

CGNet Swara’s effectiveness in securing griev-ance redress is due in large part to the moderator’s

efforts to liaise with mainstream national news jour-nalists and government ofªcials who can amplifyand act on stories reported on the system. Yet theseefforts at liaising also form the crux of the system’scomplicated relationship with journalists. Journalistswere critical of the persistence of CGNet Swara’sfounders in trying to get mainstream, nationalmedia coverage for grievances presented on the ser-vice. As one journalist questioned, “Is CGNet Swarafor the people or for journalists?” The reluctance ofthe journalists to pay attention to content fromCGNet Swara illustrates a difference of opinionbetween CGNet Swara staff and national editorsabout what is newsworthy. One journalist added thefollowing perspective:

CGNet Swara and I have the same kind of sourcenetwork. Why should I be dependent on it as asource for my news? . . . Also the content is wellsuited to local newspapers of this state than na-tional newspapers. . . . The content on it is gener-ally about NREGA and gram panchayat5 issues.There are too many of that on it. Tell me, do youthink my readership, which is urban English-speaking, is going to enjoy NREGA reports everyday? . . . My readers and my newspaper is goingto tell me that, listen—you are in a state which isthe epicenter of a very violent antistate move-ment. Give us reports about that. . . . There is noNREGA problem peculiar to this region, it can bereported from anywhere.

Another journalist said this:

We are catering to an urban readership, and ourfocus is naturally going to be on what affectstheir lives. . . . Tribals become far removed in sucha scenario. . . . When I report, I am not reportingfor the people in this state, I am reporting forpeople outside this state about this state.

A newspaper editor pointed to the capital-intensive nature of the news business to emphasizethe industry’s need to turn a proªt:

If you are investing crores6 of money you obvi-ously want proªts for it and proªts are not to befound in reporting on tribal issues. The ones whocan’t buy a newspaper and who don’t buy theproducts advertised in them—why will any news-paper pay attention to them?



5. Gram panchayats are the primary institution of local self-governance in rural India.6. A crore is a unit equal to 10 million Indian Rupees. Ten million Indian Rupees is about US$200,000.

While one journalist stated that the content wasnews although its presentation style was not,another journalist said that it was “deªnitely infor-mation, but not news. There are too many com-plaints.” Yet another journalist agreed that it couldbe a form of citizen journalism, though it seemedmore like a “discussion forum, more like anexchange of ideas and a part of citizen journalismalso because more and more people are coming upwith own ªndings and report.”

The bureaucrats and administrators have a differ-ent perspective of CGNet Swara than journalists.Referring to the limited audience of the service, onebureaucrat said that, unless CGNet Swara achievedcritical mass among the people, he would not con-sider it a serious effort in representation of people.The same administrator also suggested that a servicesuch as this would play a limited role in garneringan audience unless it scaled up to be a part of abigger entity, such as a community radio system.Concerned with the antecedents of CGNet Swara’sformation and cognizant of the region’s delicatepolitical climate, a senior bureaucrat of the state’spolice force expressed concern that the servicemight be used by people perceived to be “anti-state.” Touching on all these issues (scale, content,and perceived political slant), one ofªcial summedup his perspective:

If it [CGNet Swara] has to be taken seriously, ithas to come to some visible level and also take upvarious issues not limited to certain complaints orantigovernment issues. Then people become irri-tated and they will say that nobody is saying agood thing and only highlighting a bad thing. No-body will take it seriously if it just becomes acomplaint box.

DiscussionCoupled with the usage statistics presented earlier,the three themes we have identiªed (the experienceof users interacting with the voice interface, theemergent practice of using the platform as a mech-anism for grievance reporting, and the complex rela-tionship between CGNet Swara and establishedmedia outlets) suggest that the addition of a voiceportal to the CGNet system has both strengthenedand transformed it, although not necessarily in theways the founders would have expected.

Some of the tension appears to reside particularly

in the nomenclature of citizen journalism (Gillmor,2006). On the one hand, there are people gather-ing, submitting, and selecting stories, acting in waysthat closely resemble what journalists and editorsdo. On the other hand, the emergent practice ofusing the platform as a grievance-reporting mecha-nism suggests a use that may be unique to this sys-tem and this context, something which is not afeature common to citizen journalism platforms.

And yet, too tight a focus on nomenclature andthe deªnition of citizen journalism would detractfrom the broader lessons about the role of voice inparticipatory platforms that can be ascertained fromthe CGNet Swara case study. We frame three ofthese more generalizable issues as matters of theo-retical emphasis, rather than binary assertions.

More Participants, but a ConstrainedConversationWhile the introduction of the voice interfaceenabled more people to participate in a “conversa-tion” on CGNet Swara, the structure of that conver-sation did not necessarily improve relative to theInternet forum. If anything, since threaded two-wayconversations were more difªcult over the phone,the resulting content may have been 1) more frag-mented, and 2) more sensitive to editorial shapingby the moderators than content on the website.While it is beyond the scope of this analysis, wesuggest that two fruitful paths for further examina-tion would be to contrast text- and voice-basedforum use in resource-constrained settings in termsof 1) conversation/rhetorical analysis (turn-taking,length of statements, references to prior posts, etc.)(Hutchby, 2003); and 2) the structure of the interac-tions among users (using elements of social networkanalysis, including centrality, hierarchy, connected-ness, etc.) (Wasserman & Faust, 2004).

Text and Voice, not Text or VoiceAlthough our focus has been on users of the voicesystem, it is clear from the interviews that the sys-tem is a multiplatform endeavor; the whole remains,perhaps more now than ever, greater than the sumof its parts. That voice messages were transcribed,translated, and made available on the website was acritical part of CGNet Swara’s power to inºuenceofªcials both directly (as readers) and indirectly (viajournalists working for the mainstream press). Wewould thus suggest that, in some cases, voice mightbe a necessary but not completely sufªcient



affordance for the deployment of a participatory sys-tem in low-literacy, resource-constrained settings.Theoretical frames that account for mixed, hybrid,or complementary media may be more useful thanthose that stress successive waves of media andtechnologies. It is worth noting that, in CGNetSwara, the transferability generally ºows in onedirection, from voice to text. In the future, it wouldbe interesting to consider a text-to-voice functional-ity, whereby Web visitors could leave text commentsthat could be played in voice format to callers.

The Inºuence of Human Mediators andEditorsFinally, as is the case with many ICT4D initiatives, werisk paying too much attention to the technologyand not enough to the people and organizationsthat use the technology. In this case, the tremen-dous capability, commitment, and connections ofthe founder-moderator were absolutely critical forsetting a receptive scene for GGNet Swara in thecommunity; for conducting training workshops andmedia interviews; and for fulªlling the day-to-dayrole of selecting content, screening posts, translatingaudio to text, and championing the posts with jour-nalists and government ofªcials to effect furtherpublicity and follow-up action. It is also true that thesystem would have made little headway without thecitizen journalists being willing to use it, and with-out the mainstream journalists being willing to citeit. To switch to current ICT4D parlance, there are atleast two layers of intervention (Sambasivan, Cutrell,Toyama, & Nardi, 2010) between the individual withthe grievance and the ofªcial who can ªx it. There isboth the citizen journalist (an infomediary whoworks with information and technology on behalf ofsomeone else) generating the content, and theCGNet Swara staff (who act as moderators andgatekeepers) selecting some stories for inclusion onthe website and portal.

We think these three, sometimes overlappingthreads have value in the conversations about howvoice is integrated into a variety of current areas ofexploration around mediated social change: citizenjournalism, mentioned throughout this paper, whichis reacting to (and celebrating) the role of themobile phone as ªrst-line witness in everything fromtsunamis (Gordon, 2007) to revolutions (Comninos,2011); the ªeld of community informatics (Gurstein,

2000) which, having explored the challenges oftechnology, inclusion, participation, and socialchange for over a decade, is now embracingand innovating around the mobile phone (Baret al., 2009; Parker, Wills, & Wills, 2008); ande-governance/transparency, which has seen themobile phone emerge as a way to increase the rich-ness and range of the interactions between a stateand its citizens (Avila et al., 2010; Grönlund,Heacock, Sasaki, Hellström, & Al-Saqaf, 2010).

Each thread, at times, seeks to alter the “top-down” status quo, bringing more actors into com-plex social, political, and economic decision pro-cesses. Each is a close cousin to ICT4D, and theapproaches of each, stressing inclusion and partici-pation, are increasingly merging into a redeªnitionof how to approach ICTs in the development process(Heeks, 2010). Our integrative themes, then, mightbe to move up a level from each of the threads,addressing instead their common concerns with par-ticipation as linked to issues of power (Castells,2009) or inºuence in a public sphere (Habermas,1989).

At this level, CGNet Swara’s contributions aremixed. It is evident in our case study that the simpleinclusion of voice affordances helped to increase thereach and richness of an ongoing effort to recon-ªgure the public sphere in Chhattisgarh. Grievanceswere made salient and acted upon. To summarizethe ªve hallmarks of the project, other ICT4D pro-jects can learn from how a carefully structuredvoice-plus-text system staffed by passionate andskilled people, working with existing actors in theconventional media, both was more inclusive thanthe system that came before it and had a greaterimpact for its users and constituencies.

And yet, it is less certain that anything more fun-damental has changed (yet) because of either thepractical inclusion of voice functionality or the sym-bolic power of “voice.” This may partially be due toall the interdependencies between the new voicesystem and the existing stakeholders. Whether the“voices” were “heard” differently by “the power-ful” is perhaps a broader topic for another paper,but we suggest the work of Pettit, Salazar, andDagron (2009) and Tacchi (2008) as starting pointsfor the difªculties of untangling the symbolic andthe mundane.

Nor can a case study assess whether this same



approach would work in other settings. Would amore literate, technically savvy, Internet-connectedpopulation also elect to work with a voice platform?If not, the particular instance and promise of voiceas a means to increase a participatory project’s inclu-siveness may be restricted to low-literacy/resource-constrained settings.

ConclusionThis case study has described a participatory “citizenjournalism” system established for a place and apeople that were previously locked out of the main-stream news media. The system’s users, often (butnot always) acting as citizen journalists, share infor-mation with each other, as well as with the broadercommunity of economic and political actors in theirstate. As a system, CGNet Swara faces a constantstruggle to insert itself into the existing sociopoliticalstructures of the state of Chhattisgarh. The embed-dedness allows the system to demonstrate impactwhen ofªcials “pay attention” to the issues amp-liªed by the system and occasionally take action. It iscomplicated, and perhaps not quite what its found-ers had intended. That the system works (at all) inan environment as remote and challenging asChhattisgarh makes it a worthy case study forICT4D. And it “works” because of voice.

We describe the emerging practices aroundCGNet Swara against the background of an explo-sion of new, related practical and theoretical work inICT4D, participatory development, community infor-matics, and citizen journalism. Despite a crowdedspace, this article contributes to ICT4D theory andpractice by explicating how voice functionality canalter and enhance interactions among users, cham-pions, and traditional media—as well as those whoare most important, from the perspective of griev-ance redress—institutional actors in the manifesta-tion of “impact” and the renegotiation of power.

To the extent that CGNet Swara opened newavenues to participation (mixing the inclusivity ofthe phone/voice with the preexisting interactivity ofWeb 2.0) in a digital public sphere, and also to theextent that this participation has led to different andbetter outcomes, CGNet Swara has already hadimpact worthy of discussion and replication in ICT4Dcircles. ■

AcknowledgmentsWe are grateful to Shubhranshu Choudhary,founder and moderator of CGNet Swara, who bothdeserves principal credit for the impact described inthis article and also facilitated our ªeld work bymaking introductions to contacts in Chhattisgarh.We are also indebted to the other core members ofthe CGNet Swara team: Arjun Venkatraman, SmitaChoudhary, and Anoop Saha. CGNet Swara origi-nated from a close collaboration with Latif Alam andSaman Amarasinghe. Samujjal Purkayastha andDevadatta Sahoo also made important technicalcontributions. We are grateful to Rahul Banerjee forhelp in categorizing posts. CGNet Swara is sup-ported in part by a Knight International JournalismFellowship, under the guidance of Elisa Tinsley andBenjamin Colmery at the International Center forJournalists. Although conditions of anonymity makeit impossible to acknowledge them by name, theªeld visits would not have been possible without thesupport, hospitality, and generosity of numerouspeople in Chhattisgarh. Thanks are also due tothose CGNet Swara users, Chhattisgarh governmentofªcials, and journalists who readily granted theirtime to discuss their views and provided furtherleads to make this study possible.

ReferencesAgarwal, S. K., Kumar, A., Nanavati, A. A., & Rajput,

N. (2009, April 17–19). Content creation and dis-semination by-and-for users in rural areas. In Pro-ceedings of the Third International Conferenceon Information and Communication Technologiesfor Development, Doha, Qatar (pp. 56–65).

Avila, R., Feigenblatt, H., Heacock, R., & Heller, N.(2010). Global mapping of technology for trans-parency and accountability: New technologies.London, UK: Open Society Foundation.

Bailur, S. (2007). The complexities of communityparticipation in ICT for development projects: Thecase of “our voices.” In Proceedings of the 9thInternational Conference on Social Implications ofComputers in Developing Countries. May 2007,São Paulo, Brazil.

Banks, K., & Hersman, E. (2009, April 17–19).FrontlineSMS and Ushahidi: A demo. In Proceed-



ings of the Third International Conference on In-formation and Communication Technologies forDevelopment, Doha, Qatar (p. 484).

Bar, F., Brough, M., Costanza-Chock, S., Gonzalez,C., Wallis, C., & Garces, A. (2009). Mobile voices:A mobile, open source, popular communicationplatform for ªrst-generation immigrants inLos Angeles. “Mobile 2.0: Beyond voice?”Preconference Workshop at the InternationalCommunication Association, May 21–25. Chi-cago, IL.

Castells, M. (2009). Communication power. Oxford,UK: Oxford University Press.

Choudhary, S. (2009, June 1). CGNet and citizenjournalism in India. eJournal USA, 14(6), 42–43.Retrieved from http://www.america.gov/st/peopleplace-english/2009/June/20090616175845mlenuhret0.1840588.html

Comninos, A. (2011). Twitter revolutions and cybercrackdowns: User-generated content and socialnetworking in Arab spring and beyond. Melville,South Africa: Association for Progressive Commu-nications.

Le Dantec, C. A., Farrell, R. G., Christensen, J. E.,Bailey, M., Ellis, J. B., Kellogg, W. A., & Edwards,W. K. (2011). Publics in practice. New York, NY:ACM Press.

Donner, J. (2008). The rules of beeping: Exchangingmessages via intentional “missed calls” on mo-bile phones. Journal of Computer-Mediated Com-munication, 13, 1–22. doi:10.1111/j.1083-6101.2007.00383.x

Donner, J., Verclas, K., & Toyama, K. (2008, Decem-ber). Reºections on MobileActive 08 and theM4D landscape. In J. S. Pettersson (Ed.), Proceed-ings of the First International Conference onM4D (pp. 73–83).

Ford, M., & Botha, A. (2007). MobilED—An accessi-ble mobile learning platform for Africa. InP. Cunningham & M. Cunningham (Eds.),ISTAfrica 2007 Conference Proceedings(pp. 9–11). Maputo, Mozambique: InternationalInformation Management Corporation.

Frohlich, D. M., Bhat, R., Jones, M., Lalmas, M.,Frank, M., Rachovides, D., et al. (2009). Democ-

racy, design, and development in communitycontent creation: Lessons from the Storybankproject. Information Technologies & InternationalDevelopment, 5(4), 19–35.

Gillmor, D. (2006). We the media: Grassroots jour-nalism by the people, for the people. Sebastopol,CA: O’Reilly.

Goggin, G., & Clark, J. (2009). Mobile phones andcommunity development: A contact zone be-tween media and citizenship. Development inPractice, 19, 585–597.

Gordon, J. (2007). The mobile phone and the publicsphere: Mobile phone usage in three critical situ-ations. Convergence: The International Journal ofResearch into New Media Technologies, 13(3),307–319.

Grönlund, A., Heacock, R., Sasaki, D., Hellström, J.,& Al-Saqaf, W. (2010). Increasing transparencyand ªghting corruption through ICT: Empoweringpeople and communities. Stockholm, Sweden:The Swedish Program for ICT in Developing Re-gions (SPIDER).

Gurstein, M. (2000). Community informatics:Enabling communities with information and com-munications technologies. Hershey, PA: IdeaGroup.

Habermas, J. (1989). The structural transformationof the public sphere: An inquiry into a categoryof bourgeois society. Cambridge, MA: MIT Press.

Heeks, R. (2010). Development 2.0: NewICT-enabled development models and impacts.Communications of the ACM, 53(4), 22–24.doi:10.1145/1721654.1721665

Hutchby, I. (2003). Affordances and the analysis oftechnologically mediated interaction: A responseto Brian Rappert. Journal of the British Sociologi-cal Association, 37(3), 581–589.

India Ministry of Communications & InformationTechnology. (2010). Broadband services.Retrieved from http://pib.nic.in/newsite/erelease.aspx?relid�71124

Koradia, Z., & Seth, A. (2012). PhonePeti: Exploringthe role of an answering machine system in acommunity radio station in India. In Proceedingsof the Fifth International Conference on Informa-



tion and Communication Technology for Devel-opment, Atlanta, GA (pp. 278–288).doi:10.1145/2160673.2160708

Kotkar, P., Thies, W., & Amarasinghe, S. (2008,April). An audio wiki for publishing user-gener-ated content in the developing world. HCI forCommunity and International DevelopmentWorkshop, CHI 2008 (pp. 1–2).

The Manthan Award. (2008). Manthan AwardSouth Asia winners. Retrieved from http://manthanaward.org/section_full_story.asp?id�665

Midgley, J., Hall, A., Hardiman, M., & Narine, D.(1986). Community participation, social develop-ment, and the state. New York, NY: Methuen.

Ochoa, X., & Duval, E. (2008, April 22). Quantitativeanalysis of user-generated content on the web.Proceedings of WebEvolve 2008 Web ScienceWorkshop at WWW2008 (pp. 19–26).

Odero, B., Omwenga, B., Masita-Mwangi, M.,Githinji, P., & Ledlie, J. (2010, December 17–18).Tangaza: Frugal group messaging through speechand text. First Annual Symposium on Computingfor Development, London, UK.

Parker, M., Wills, G., & Wills, J. (2008). Communityin tension (CiT). (Technical Report ECSTR-LSL08-002, ECS). Southampton, UK: University ofSouthampton.

Patel, N. (2010). Designing and evaluating voice-based virtual communities. Proceedings of CHIExtended Abstracts, ACM (pp. 2963–2966) April10–15.

Patel, N., Chittamuru, D., Jain, A., Dave, P., & Parikh,T. S. (2010). Avaaj Otalo—A ªeld study of an in-teractive voice forum for small farmers in rural In-dia. ACM CHI (pp. 733–742), April 10–15.doi:10.1145/1753326.1753434

Pettit, J., Salazar, J. F., & Dagron, A. G. (Eds.).(2009). Introduction to citizens’ media and com-

munication [Special Issue]. Development in Prac-tice, 19, 443–452.

Plauché, M., & Nallasamy, U. (2007). Speech inter-faces for equitable access to information technol-ogy. Information Technologies & InternationalDevelopment, 4(1), 69–86.

Ranganathan, K., & Sarin, A. (2011). A voice for thevoiceless: Peer-to-peer mobile phone networksfor a community radio service. Ahmedabad, In-dia: Indian Institute of Management. Retrievedfrom http://www.iimahd.ernet.in:8181/assets/snippets/workingpaperpdf/15380481872011-05-02.pdf

Sambasivan, N., Cutrell, E., Toyama, K., & Nardi, B.(2010). Intermediated technology use in develop-ing communities. Proceedings of the 28th Inter-national Conference Extended Abstracts onHuman Factors in Computing Systems, CHI’10,ACM, (p. 2583), April 10–15.

Sherwani, J., Ali, N., Mirza, S., Fatma, A., Memon,M., Karim, M., et al. (2007). HealthLine: Speech-based access to health information by low-literate users. In Proceedings of the Second Inter-national Conference on Information and Commu-nication Technologies for Development (pp. 1–9).

Slater, D., Tacchi, J., & Lewis, P. A. (2002).Ethnographic monitoring and evaluation of com-munity multimedia centres: A study of Kothmalecommunity radio Internet project, Sri Lanka. Lon-don, UK: Department for International Develop-ment.

Sterling, S. R. (2009). Advancement through interac-tive radio. Information Systems Frontiers, 11(11),145–154.

Tacchi, J. A. (2008). Voice and poverty. Media Devel-opment, 55(1), 12–15.

Wasserman, S., & Faust, K. (1994). Social networkanalysis: Methods and applications. Cambridge,UK: Cambridge University Press.



Healthcare

CGNET SWARA WP