Reconciliation and the Web

Today the World Wide Web has become a singular method for information dissemination and interaction. So it comes as no surprise that truth and reconciliation commissions [TRCs] have increasingly turned to the web to help them publicize their work, disseminate their findings, and, to a lesser degree, interact with their constituencies. Indeed, today’s online interactive, social and rich multimedia facilities (sometimes referred to as Web 2.0) seem deeply suitable not only to helping a truth commission distribute information, but also as a tool to assist the core missions of a TRC, that of providing a platform for victims to tell their stories and an opportunity for all affected parties to interact and engage with each other in healing dialog.^[1]

Through a systematic study of Liberians using these technologies under controlled experimental conditions, we found evidence of the ability of interactive and rich multimedia, such as audio and video material, to effect post-conflict psychological healing.^[2] Our hypothesis is that TRC websites that support interaction and rich multimedia offer a powerful environment for facilitating post-conflict reconciliation and healing.

This paper offers a case study in the design, development, and operation of an interactive multimedia website for Liberia’s Truth and Reconciliation Commission. This has been a project of the Technologies and International Development Lab at the Georgia Institute of Technology. After an overview of the use of websites by different truth commissions, the paper describes the process by which we designed a new website for Liberia’s truth commission that became a process-oriented, interactive multimedia site. We will then examine the patterns of use observed on the site and end with our conclusions and observations for future work.

TRCs and the World Wide Web

Truth commissions have been slow to adopt the internet as a central tool to their work. In a 2007 analysis of TRC websites, we discovered that some commissions had successfully used the internet to disseminate reports and information, but none were making full use of the web’s interactive features or employing rich media such as video and audio.^[3] This study was based on content analysis of nine truth commission websites. The websites were chosen through an exhaustive search of official truth commission sites across the web. Unofficial websites, such as those maintained by international organizations, were excluded from the study. At least two websites, including one from Guatemala,^[4] were discarded because of their poor information value, and some websites were excluded due to language issues (e.g., a site from South Korea^[5]). This resulted in a study set of nine commission websites, representing East Timor, Ghana, Greensboro (United States), Haiti, Liberia (the original commission site prior to this study), Peru, Rwanda, Sierra Leone and South Africa. Results of the study are summarized in Table 1 below; all sites were accessed in July of 2007.

Each website was analyzed for its complete feature set. While dissemination of a commission’s final report seems like a base facility to offer on a TRC website, only four of the eight study sites (excluding Liberia, which at the time had not released a final report) actually provided a link to their report. For example, East Timor did not update its site to include this document despite the fact that the report was delivered on July 2005 and their main page states, “We will post information about the final report and these products on this website soon, including details of how you can order a copy of the report.”

Many of the websites had sections devoted to media reports, from which they provide access to press releases, articles from print, and broadcast media. In fact, seven of the nine analyzed sites included some kind of media related content. Another common feature of these sites was the presentation of information about the members of the commissions. Six of the websites included pictures and biographies of the commissioners, and in the case of Sierra Leone, personal email addresses were provided so that interested folks could contact them. Most sites provided ways to contact the entire commission via email and indeed only two of the websites (Haiti and Peru) did not provide some way of contacting the commission via email. Five of the websites include related links to local or international institutions and other TRC sites.

More advanced features, such as search capabilities and multimedia files, were less common. Only Peru, Sierra Leone and East Timor include search capabilities, that is, a box on the site itself that allowed users to search keywords within the site pages. Only the Peru and Greensboro sites offered some form of video content. In the case of Peru, the website offered video testimonies of people who told their stories to the commission. Greensboro’s website included videos of the swearing-in ceremony of the commissioners as well as an interview with its research director. In addition, during hearings in 2005, the website provided live coverage via streamed video. Other sites included transcripts, though no videos, from hearings (East Timor, Sierra Leone, Peru, and Greensboro). While video content was rare, six of the websites published pictures from commission hearings, inaugural events, and other meetings.

In summary all sites were designed in some ways to disseminate and inform by sharing information about the commissioners or to distribute electronic copies of their final report (Rwanda was a bit of an exception in not providing any of these features). Most sites also provided some means to connect by offering contact information and email addresses for the commission. In our analysis we found that no sites offered facilities to interact, such as discussion fora, places to upload images, or ways to comment publicly on commission material. And, finally, only a couple sites offered much in the way of multimedia features such as video or sound.

TRC of Liberia Website Design

In early 2007 after a series of meetings with Jerome Verdier, Chair of Liberia’s Truth and Reconciliation Commission, and other commissioners and stakeholders, the Technologies and International Development Lab at the Georgia Institute of Technology undertook a collaborative and participatory design exercise with the aim of redesigning, implementing, and hosting the Liberia TRC’s website. While the commission had already developed a simple web presence (Figure 1), there was a shared recognition that the site was failing to meet all of the possibilities of an interactive, online presence. As noted in our analysis from the previous section, while the original Liberia website had capacities to inform and to connect, it lacked facilities to interact and did not make use of multimedia materials.

Figure 1. Original Liberia TRC website prior to our re-design efforts

To this day, Liberia has very limited internet penetration and, indeed, low levels of computer and print literacy among its population.^[7] Thus it was immediately evident that, given these practical constraints, the TRC website user base would consist primarily of a small elite in Liberia, Liberians in the Diaspora, and a broad base of other international and domestic stakeholders of the post-conflict reconciliation process, such as relevant NGO’s, international organizations, and so forth. While we recognized this broader potential user base, based on our consultations with the TRC commissioners, we identified Liberians in the Diaspora as our most important user population, and therefore we set out to design a site that would focus mostly on their needs and interests.

In order to redesign the TRC online portal, indentifying this principal end-user population was our critical first step.^[8] We then employed participatory design methods with representatives from this end-user group—in particular Liberian expatriates from the Atlanta area—and held consultations with TRC officials. On one occasion we gathered advice from visiting Liberian dignitaries.^[9]  This is a component of our Heuristic-Diaspora-Field design method.^[10] In total, eight meetings took place with this set of informants to cover the requirements-gathering, user analysis, and initial design evaluation phases.^[11] This series of meetings was begun in the first quarter of 2007. Based on information gathered from these meetings, we formulated an initial list of requirements for the site that included:

• Support for online entry of formal written statements to the commission;
• Repository of multimedia content resulting from TRC proceedings;
• Communication channels with the TRC;
• Moderated discussion forum and collaboration space for discussion and interaction;
• Secure, encrypted storage of submitted official statements and prominently displayed security guarantee;
• Linkages with other TRCs as part of the global TRC movement.

Figure 2. Some preliminary redesigns employing original site's structure and imagery 

During this initial set of meetings, we brought to our user group some very preliminary design sketches and explored new imagery and color palates while still maintaining the original site’s organizational structure (Figure 2). These early design sketches were used to facilitate dialog and elicit reaction and response from our end-user informants. This process quickly established an important, over-arching design principle: namely, our expatriate informants stressed that the original TRC Web site (and our initial redesigns) overly relied on the bureaucratic structures of the commission itself as their organizing principle. The site was organized around the offices of the commissioners, the various commission departments, and so forth (this Weberian structure is, we find, the standard for most institutional sites). In contrast, our end-user informants argued that the TRC site should instead be structured around the TRC’s reconciliation process and the various constituent goals of this process.

Following on these focus group findings, the next site mock-ups we presented were structured around the core set of reconciliation processes and goals we identified in the user meetings: speaking the truth, forgiveness, reconciliation, and justice. Figure 3 shows a preliminary design mock-up featuring this process orientation; the four goals are presented on the left-hand-side of the page under the heading “Our Process.”

Figure 3. First design mock-up to feature a reconciliation process orientation

The focus groups also revealed that the choice of the site’s visual imagery would be both critically important and difficult. While visually conveying the ideals and “Liberian-ness” of the TRC was essential, finding appropriate and broadly acceptable images was difficult given Liberia’s ethnic and linguistic diversity. Avoiding bias toward any one such group was critical. These considerations affected choice of color scheme, visual icons, and other decorative imagery.

For instance, when we asked “What is Liberian,” our user group members agreed only on the Liberian flag. Based on that, some suggested we use the flag colors of red, white, and blue for the site while others felt that this would make the TRC, an independent commission, appear too closely aligned with the government. Still others pointed out that in traditional cultures red was commonly associated with blood, war, and violence.

Based on these at times conflicting sets of user inputs, we developed a new design mock-up that implemented both the process orientation and inputs on color and imagery (Figure 4). This new design had an overall green palate, which was seen as neutral and comforting. The TRC logo was deemed an important branding element so it was included, and the Liberian coastline was integrated into the head banner. This coastline represented Liberia to our users without exercising any difficult politic or narrow identity. The reconciliation process was encoded in the original four steps with accompanying iconography:

• Speak the Truth, represented by a palava hut;
• Forgive With an Open Heart, represented by the sharing of cola nuts;
• Justice, represented by scales;
• And Reconcile, represented by two clasped hands.

The front page also had facilities for a slide show, news items, and navigation facilities.

Figure 4. First design mock-up, implementing the full set of inputs from the user participatory design meetings, including process orientation and new color palate and esthetic

In addition to the Liberian coastline as an element in the title banner, the initial design included a photograph of the white Strophanthus Gratus flower, selected based on the advice of a botanist at New York’s Museum of Natural History who claimed that the species was endemic to Liberia (Figure 4); however, at a subsequent focus group meeting, the choice of the flower was rejected by our Liberian informants who claimed to have never before seen such a flower. This finding demonstrates how domain experts, such as the botanist, can never fully represent actual user populations. In the end, the flowers were replaced with a familiar image of palm fronds.

Another result of the focus group engagements was a call for the inclusion of several interactive multimedia content areas, including photo, audio, and video galleries. Other areas for interaction were added including discussion fora, means to contact commissioners, and a facility to offer sworn, sealed testimony over the Internet. This would be the first time that a truth commission had accepted sworn official testimony over the internet. Finally some of the more “traditional” TRC web components were kept, including areas for reports, details on the commissioners, the TRC schedule and programs, and so forth.

In addition, based on user inputs, a fifth process element was added to the four already in place, highlighting web material that explained how the site was safe and secure.  A skeleton key icon represented this concept. This was an important element, especially given that the site allowed for direct user contributions and included an area where visitors could provide sworn testimony.

Figure 5. Final design with process elements now in an oval, the inclusion of the safe and secure icon, and the removal of the justice icon.

The final step prior to launching the new site was to work with the TRC commissioners and chair to ensure that the design met with their approval. These interactions resulted in one significant change, namely the removal of the Justice process icon. Cllr. Verdier noted, quite correctly, that the TRC’s mandate focused on transitional justice measures but stopped short of formal justice. Thus the inclusion of justice as an element of their process was inaccurate and inappropriate. Figure 5 shows the final front page for the site, which was finalized, tested, and launched in October 2007.

Usage Patterns of the Website

The new site was launched with special instrumentation to record all visits to the site and accumulate various points of data on each visit, such as exactly which pages were visited, how long a user stayed on a page, and where the visitor was located, etc.^[12] This data allows for evaluation of the reach and impact of the site. A simple figure of merit for the website’s impact was the number of unique visitors the site attracted. The web visit analysis tool distinguished between the raw total number of visits, which counts repeat visits from the same person as separate impressions, versus the number of unique visits, which essentially counts the number of individual people who visited the site. Figure 6 graphs the number of unique visits per day from the initial site launch until August of 2011.

The histogram shows some interesting dynamics in the number of unique visits to the website over its initial years. At the left of the graph, during the month of March in 2008, we see a period logging zero visits. In point of fact, during this period of time there were visits to the site. However, due to an error in our logging tool installation, all visits during this month were lost. Ignoring this short period without any visitor data, on average the site received nearly 2,500 unique visitors per month. Starting in mid-2009 there is a visible, steady increase in monthly visits to the website. This continuous growth in interest spikes with the release of the preliminary final report in July of 2009 and spikes again with the release of the final report at the end of the year. Indeed during the month of this final report release, November 2009, the site experienced 10,935 unique visits. The number of visits then fairly quickly drops off following the final report release. Certainly this patterning makes sense: as public hearing activity builds and anticipation of the final report grows, we see the steady rise in visits culminating in a dramatic number of visits for the final report and then a fairly rapid decline back to a somewhat steady state of 1,000 unique visits a month.

 Figure 6. Histogram of over three years of number of unique visits per day to website

Beyond the raw number of unique visitors to the website, the visitor logging tool saves some information about each visit, including a rough attempt to locate each visitor geographically. This process is not precise and, in particular, many visitors from Liberia (and other African countries) will be logged as coming to the site from other locations. With that caveat in mind, Figure 7 shows the estimated percentage of visits by sub-continental region. Nearly 80 percent of visits are logged from North America. Northwest Europe then accounts for close to 11 percent of visits, but recall that visits from Liberia may be logged as coming from North America or Europe. Next is West Africa with 4 percent. This is understandably the largest region of Africa, with South Africa accounting for less than 1 percent and East Africa even smaller. Western Asia, more commonly called the Middle East, account for about 2 percent of visits.

Technical challenges to these visit logs notwithstanding, it seems clear that the large majority of visitors come from North America, and indeed specifically the United States, forty times more visits than Canada. When this data is broken down by U.S. city, the most significant traffic comes from cities with large Liberian Diaspora, 17 percent originating in New York City alone, 3 percent from Miami, 3 percent from Minneapolis, 2 percent from Chicago, and 1.5 percent from Washington DC.

These results seem to confirm our initial design principle—that the website would be visited mostly by people in countries with substantial internet infrastructure and penetration, such as North America and Europe, and Liberian use would come most significantly from the Diaspora.

Figure 7. Percentage of visits by sub-continental region

Finally, the traffic analysis not only counts total visits to the website as a whole but also isolates the number of visits to each part of the website, counting the number of visits for each. This data highlights those parts of the site that were of the most interest to visitors. Table 2 lists the twenty most visited areas of the website; these are either specific pages or groups of pages that are coherently related. By far the most visited areas of the site were the video homepages and video search areas. These locations are not the pages with specific videos but are instead the homepage and navigational and search facilities where visitors are able to locate specific videos they are interested in viewing. The site contains over 1,000 videos available for playback, nearly all of which document the series of public hearings conducted by the commission. While the general video navigation facilities are the most visited part of the site, the eighth most visited page (with 1.1 percent of all page views) is the video of the hearing of Prince Johnson. Prince Johnson was a notorious rebel force commander during the civil war and is currently a politician in Liberia. Beyond Prince Johnson, other specific videos also enjoyed significant viewership. For instance, the fifteenth most visited portion of the site contained videos from the Diaspora hearings held in the USA.

Beyond videos, the second most popular page on the website was the homepage itself (www.trcofliberia.org, depicted in Figure 5 above) with 13.4 percent of all page views. The third most popular area of the website was the final report section where visitors could download the report material. This area enjoyed 4.4 percent of all visits to the website.

In summary, Table 2 underlines how popular, and we hope impactful, the rich media material of the site was. Videos were by far the most popular materials. While these facilities focus on informing visitors on the activities of the commission, in other work we have demonstrated that viewing rich media can be effective in processes of national reconciliation and healing.^[13] The second most popular facility on the site was information delivery of primarily text materials such as the final report, news, information about commissioners, and so on. The third most popular type of facility is the interface component that connects visitors to the commission and allows them to interact. The interaction facility to submit formal statements accounted for 0.8 percent of all visits to the website. And 0.5 percent of visits were made to the page with information on how to contact the commission itself. However, the interaction pages allowing visitors to contribute public material such as pictures or the discussion fora of the website did not receive many visits at all. So while some elements of connection and interactivity were appreciated, they enjoyed considerably less traffic compared with the rich media facilities and the more traditional informational pages.

As the TRC has wound down its operation, we have transitioned the site to a static one that sits now on an external web host company and will be there for ten years or so.
 

1. 1. The author is deeply grateful to the Truth and Reconciliation Commission of Liberia for its trust, collaborative sprit and friendship. In particular we thank the Chair, Cllr. Jerome Verider. In addition we thank the Executive Director Mr. Nathaniel Kwabo, and the IT specialist Mr. Dickson Fully. Design, implementation and support for the website was the output of a large team of committed Georgia Tech students. In particular we thank Martin Bednar, Stephanie Emling, Corbin Le Pon, Parker McGee, Daniel Serrano-Baquero, and Thomas Smyth. Arkadiusz Banasik was the site’s principal design professional. The Georgia Tech College of Computing computer support staff was instrumental in keeping the site available for over three years. In particular we thank Michael Luttrel, Peter Wan, and Keith Watson. As our Liberia based project director, John Etherton provided constant assistance and advice. The Carter Center has been a constant friend to our Liberia projects, and we in particular we thank Thomas Crick and Itonde Kakoma. William Long was Chair of the Sam Nunn School of International Affairs throughout this project and we are profoundly grateful for his constant support. The Atlanta based Liberian Diaspora community welcomed this project and provided instrumental design inputs. We thank Mina and the customers of Mina’s Kitchen in Clarkston, Georgia, and we thank the many members of the Liberian Association of Metropolitan Atlanta. Finally, we thank the people of Liberia, those within her shores along with those in the Diaspora.
2. 2. Michael L. Best, William J. Long, John Etherton, and Thomas Smyth, “Rich Digital Media as a Tool in Post-Conflict Truth and Reconciliation,” Media, War and Conflict 4, no. 3 (December 2011): 231-249.
3. 3. J. A. Muñoz, The Use of Internet in Truth and Reconciliation Commissions: Missed e-opportunity? (Atlanta, GA: Georgia Institute of Technology, Sam Nunn School of International Affairs, Atlanta, 2007).
4. 4. The Office of Human Rights of the Archbishop of Guatemala, http://www.odhag.org.gt.
5. 5. See http://truthfinder.gov.kr/.
6. 6. Original website prior to redesign.
7. 7. The World Bank, “World Development Indicators (WDI) 2010,” http://data.worldbank.org/data-catalog/world-development-indicators/wdi-2010.
8. 8. Hugh Beyer and Karen Holtzblatt, Contextual Design: Defining Customer-Centered Systems (Morgan Kaufmann, 1998).
9. 9. Douglas Schuler and Aki Namioka, Participatory Designs: Principles and Practices (CRC/Lawrence Erlbaum Associates, 1993).
10. 10. Thomas N. Smyth, John Etherton, and Michael L. Best, “MOSES: Exploring New Ground in Media and Post-conflict Reconciliation,” in Proceedings of the 28th International Conference on Human Factors in Computing Systems (New York: Association for Computing Machinery, 2010), 1059–1068.
11. 11. Michael L. Best, Thomas N. Smyth, Daniel Serrano-Baquero, and John Etherton, "Designing for and with Diaspora: A Case Study of Work for the Truth and Reconciliation Commission of Liberia,” in Proceedings of the 27th International Conference Extended Abstracts on Human Factors in Computing Systems (New York, Association for Computing Association Machinery, 2009), 2903–2918.
12. 12. We used the Google Analytics technology, http://www.google.com/analytics/. 
13. 13. Michael L. Best, William J. Long, John Etherton, and Thomas Smyth, “Rich Digital Media as a Tool in Post-Conflict Truth and Reconciliation,” Media, War and Conflict 4, no. 3 (December 2011).