What do you remember about the 1984 gas leak at the Union Carbide pesticide plant in Bhopal, India? My guess is that if you do not know much about it, you would probably Google it. The first search result would most likely be the "Bhopal disaster" Wikipedia article . There you would find linked citations to information about the causes of the gas leak, the previous warnings and other incidents, the health effects from gas exposure, the legal actions against Union Carbide, the ongoing water contamination in Bhopal, and the Yes Men's settlement-fund hoax on BBC News.
Although the Bhopal incident happened over a quarter century ago, its heritage is still alive, mainly due to the growing digital presence of memories that capture different perspectives of the event across multiple generations. The contributors to the "Bhopal disaster" Wikipedia article, for example, include a medical doctor, a chemical engineer, a chemist, an Indian native, and a disaster enthusiast who enjoys editing relevant Wikipedia articles. With the advent of social media and other networked technology, the heritage of historic crises like the Bhopal gas leak are being kept alive and continually curated in socially distributed ways.
A new kind of heritage is being born, as the general public increasingly use social media to document their experiences online. This civic ritual of using new media to share stories at nearly global scales is not only culturally significant but also ripe with design opportunities. Through advances in information and communication technology (ICT) and the rising use of social media as one class of ICT, we are on the edge of fundamental changes in how heritage and the memories it encompasses are being produced, preserved, interpreted, and shared online .
The crisis domain provides a unique opportunity to look at heritage issues across time, space, and networks of people. Major crises and disasters tend to generate collective memories and varying narratives. They also tend to amplify our preexisting vulnerabilities and test our resilience to disasters. To some degree, how we remember and learn from historic crises can directly affect their existence and application to present-day society and current cultural practices. The key is determining what kinds of memories will strengthen our resilience to potential crises when shared with present and future generations.
One of my approaches to investigating this new form of heritage in the networked world involved looking at heritage in terms of its form and content. Content encompasses the kinds of narratives and themes emerging from collective memories. Form refers to the medium in which these memories exist. What emerges in examining both is a practice of curation. There is a craft involved with curating these memories in a way that shapes our heritage collectively.
To some extent, social media services support tasks associated with curation (e.g., collecting, organizing, storytelling, and exhibiting) by design. Social media also supports participatory forms of curation by encouraging distributed ways of preserving memories and making sense of their heritage. Designers of social technology now have the opportunity and perhaps the responsibility to decide how memories are kept aliveor curatedthrough the values they embed in the design of their technology.
As a way of opening the design space for heritage matters, I embarked on a research project that investigated the sociotechnical practices of using social media platforms and services to curate the narratives of historic crises . After surveying the social media presence of 111 major crises that occurred between 1960 and 2010, I discovered trends in the types of hazards and disasters that already had a rich heritage in the social media world, themes in the kinds of narratives emerging across different historic crises, and patterns in the use of social media services that reflected different types of curatorial activities. I then conducted virtual ethnographic investigations to uncover the trending narratives across four distinct crisis events: the 1984 Bhopal gas leak, the September 11 attacks, 2005's Hurricane Katrina, and the climate change crisis.
I conducted a final study that systematically investigated the concept of curation and the diverse set of curatorial activities it encompasses. This led to the development of a theoretical model called socially distributed curation that provides indepth descriptions of how curation encompasses a set of activities (e.g., collecting, categorizing, verifying, synthesizing, exhibiting, guiding, and maintaining) and how social media features facilitate these curatorial activities in distributed ways. What I have discovered are budding examples of living heritage emerging in the social media landscape that can provide inspirations for how we design and use ICT to support heritage matters in new ways.
The 1984 Bhopal gas leak is an example of a disaster that already has a rich living heritage in the social media landscape. There are more than 200 Facebook groups, pages, and causes; more than 300 Flickr images; more than 1,000 Delicious bookmarks; more than 88,000 blog posts; and more than 600 YouTube videos pertaining to the Bhopal disaster. There are also more than 1,000 editors for the "Bhopal disaster" Wikipedia article. The heritage of the Bhopal disaster is thriving as people share memories of what happened in 1984 and updates on the impact of the Bhopal disaster in the present day.
What is fascinating about the story of the Bhopal disaster in the networked world is the range of stories that reflect its complex history. We are moving away from brief textual accounts of these disasters reiterating key dates and numbers, such as the following snippet from the Bhopal entry on Encyclopedia.com:
"In Dec., 1984, a cloud of methylisocynate (MIC) gas escaped from the Union Carbide plant in Bhopal. An estimated 3,000 to 7,000 died immediately, 15,000 to 20,000 died from the effects in the years after the disaster, and 50,000 to 100,000 suffered from serious injuries as a result of the world's worst chemical disaster. The Indian government sued on behalf of 570,000 victims and in 1989 settled for $470 million in damages and exempted company employees from criminal prosecution. The Indian judiciary rejected that exemption in 1991, and the company's Indian assets were seized (1992) after its officials failed to appear to face charges" .
What we are moving toward is a more comprehensive and collaborative account of these disasters, distributed across multiple forms of media. The heritage of the Bhopal gas leak has transformed into a living heritage of the Bhopal disaster at a larger scale. It encompasses a history going as far back as the Green Revolution in the 1970s, which caused an increase in agriculture and pesticide production. It also includes stories about the ongoing water contamination from the pesticides that are reportedly still left at the Union Carbide (UC) pesticide plant in Bhopal. The value and meaning that come from remembering the heritage of historic crises like the Bhopal gas leak emerge from understanding the impact of the disaster in the present day, why it occurred, and, more important, how we can be prepared and engage in disaster mitigation to prevent today's crises from becoming disasters and catastrophes in the future.
In contrast to Union Carbide's interpretation of what caused the gas leak (employee sabotage), the "Bhopal disaster" Wikipedia article presents an extensive list of causes that many believe contributed to the gas leak. The "Contributing factors" section of the article states, "The use of a more dangerous pesticide manufacturing method, large-scale MIC storage, plant location close to a densely populated area, undersized safety devices, and the dependence on manual operations" together caused the gas leak or at least exacerbated it. Bhopali residents were not informed in advance what protective actions to take or how warnings would be communicated. The article further cites plant-management deficiencies, including a lack of skilled operators, the scaling back of safety management, poor maintenance, and inadequate emergency action plans. The true test of our resilience to a crisis like this is whether or not we learn from such causes documented by past and present generations and then make social, technical, and political changes to prevent them from happening in the future.
What also emerged in the stories from and about the survivors of the Bhopal disaster resurfacing in Facebook groups, YouTube videos, and blog posts were narratives explaining the protective actions that were taken or could have been taken during the gas leak (e.g., evacuating to the nearby lake or covering the face with a wet cloth since the gas is soluble in water). In the Students for Bhopal Facebook group, Bhavna Suri from New Delhi shares the memories of a Bhopali nativea teenager at the time of the gas leak who had a close relative working at the Union Carbide plant. That relative died while trying to fix the leak. In the group's discussion section she goes on to explain the lack of risk communication on the day of the gas leak:
"The knowledge that a disaster was going to happen was known to UC officials earlier in the evening, at about 5 p.m. The gas leaked into the city around 11 p.m., when most of the people in the immediate area of the UC plant were at home. None of them were warned about the potential disaster. The only people who supposedly knew were politicians and government officials, who had fled from the vicinity. This happened in the era when there were no cell phones, text messaging, or email and there was only one sleepy government-run TV channel in India. Was the impact on the rest of the world lesser because of it? Absolutely!"
Suri's post in this Facebook group not only offers an interpretation of the risk communication issues that emerged from the Bhopal gas leak at that time, but also reflects on the significance of these issues in the present day with the advent of pervasive ICTs being used in the crisis domain. If the Bhopal gas leak happened in today's era of digital and networked communications, would the response have been different? Trends in social media use during recent disasters (e.g., Arab Spring, the 2010 Haiti earthquake, and the 2011 Japan earthquake) provide a glimpse into how these networked technologies can play a critical role in communicating timely and actionable information that could potentially save lives.
The act of keeping the memory of the Bhopal disaster alive is a subtle yet powerful form of action that is critical to maintaining the living heritage of this ongoing crisis. Passing on the memories from the past can enable action-oriented narratives that help raise awareness and promote disaster mitigation to strengthen community resilience to future crises. At the same time, solicitations for and documentation of direct actions via social media platforms are increasing, since these platforms enable mass coordination and mobilization at a global scale. For example, advocacy organizations use Facebook groups to send action alerts to group members, urging people to engage in activities related to Bhopal that need immediate attention and action. Many of the Bhopal-related photos on Flickr tend to be images of die-in protests, hunger strikes, and torch parades happening in India and around the world.
Although these direct actions typically occur out in the real world, they are largely organized and publicized using social media. What networked technologies allow is a new kind of action that can take place remotely in socially distributed ways. The Students for Bhopal organization, in collaboration with the International Campaign for Justice in Bhopal (ICJB), has been using social technology for the past four years to mobilize an international social justice movement to connect communities worldwide that are actively involved in a variety of issues related to the Bhopal disaster (e.g., chemical safety, corporate accountability, and environmental justice).
These narratives extend long after, as well as long before, the gas leak in 1984, and therefore illustrate the dynamic and continuous qualities of heritage-related activity. The existence of these action-oriented narratives alone points to the present-day impacts of the crisis, plus the strong need to keep the heritage of the Bhopal disaster alive for the benefit of present and future generations. As people increasingly use ICTs to document and curate the narratives that emerge from historically significant events, we must consider how we design technologies, policies, and practices that enable the sharing of digital artifacts, not only for those who can immediately harness their value for disaster-mitigation purposes, but also for our descendants to strengthen their resilience to future crises.
We are beginning to see how disaster narratives are made visible and kept alive in the digital world through the remixing, reinterpretation, and representation of memories using new media. The heritage practices around historic crises that are beginning to emerge from the use of social technology are fruitful grounds for designing sociotechnical systems that can transform how we view preservation and curation in the networked world.
Socially distributed curation is a sociotechnical practice involving people, cultural artifacts, and information and communication technology. This type of curation is collaborative and distributed, creating shared ownership over the stewardship of the living heritage being sustained. It is a participatory, bottom-up approach that makes each curatorial activity transparent to allow other interested parties to partake in the curatorial process.
It also entails a spectrum of micro and macro forms of curation in the digital world. On one side of the spectrum, one might view the Web as a constantly growing archive, with Google often seen as the ultimate curator. On the other side of the spectrum, Flickr group members engage in micro-curation by collectively organizing each other's photos into thematic groups. Curation also happens more quickly today through harnessing the networking capacities of the Internet. Therefore, curation is not only practiced as a long-term endeavor, but also is increasingly engaged through real-time activities. Future information systems could be designed to facilitate the curation of both deep histories that preserve multiple generations of heritage and minute histories that capture and represent memories on specific topics in real time.
What emerged in my analysis of curatorial activities was a spectrum between human and algorithmic forms of curation. Human-aided algorithmic curation  involves human discernment in selecting and filtering artifacts that were initially aggregated algorithmically. ICTs can be designed to help users manually filter, organize, and manage content. Algorithmic-aided human curation, such as the PageRank algorithm, relies more on algorithms to aggregate and filter content but is still based on human actions. Algorithms are increasingly augmenting human-aided curatorial activities in the social Web context. For instance, social media users tend to select and view content based on the number of views, ratings, friends, followers, and/or subscribers it has, all of which can be calculated using algorithms and then become algorithmic recommendations. Based on my investigation of the concept of curation, it is better to allow algorithmic forms of curation on the input side and not rely solely on algorithms when curating on the output side.
As interaction designers, we must consider when it is better to use algorithms or human discernment to help us curate these ever-flowing information streams from the social media landscape. Algorithms are becoming powerful mechanisms to manage and represent the influx of online activities, but it is important to be aware of how these algorithms affect what gets valued and devalued. It is tempting to believe the algorithms that power many Web services we use today are democratic in their calculations, but we need to be critical of them, too, and understand the values embedded in the design of these social networking platforms and other ICTs.
There is an opportunity here to begin exploring how we might preserve a deeper and richer heritage of historical events using large-scale information systems and networked technologies. Preserving generations of memories that appear in multiple forms of media will require curatorial tools for networks of people. Computational methods and machine-learning techniques will need to be developed to more effectively curate large datasets and maintain data provenance when collecting streams of data from multiple media platforms over long periods of time. We must also consider how to design ICTs that guide rather than control human discernment when curating collective memories. We are just beginning to see the potential of sustaining heritage in new ways across space, time, and networks. It is an opportunity in which we should all participate, because it shapes who we are and what we value as a society and as a generation.
2. Liu, S.B., Palen, L. and Giaccardi, E. Heritage matters in crisis informatics: How information and communication technology can support legacies of crisis events. In Crisis Information Management: Communication and Technologies, C. Hagar, ed. Woodhead Publishing Limited, Cambridge, U.K., 2011.
3. Liu, S.B. Grassroots heritage: A multi-method investigation of how social media sustain the living heritage of historic crises. Ph.D. Dissertation, University of Colorado at Boulder. Ann Arbor: ProQuest/UMI, 2011. (Publication No. AAT 3453824.)
5. Human-aided algorithmic curation is a term that journalist Jeff Jarvis used in a blogpost entitled "Content farms v. curating farmers"; http://www.buzzmachine.com/2009/12/14/content-farms-v-curating-farmers/
Sophia B. Liu conducted this research as an NSF Graduate Research Fellow at the University of Colorado at Boulder in the Technology, Media and Society interdisciplinary Ph.D. program at the ATLAS Institute. She is currently a Mendenhall Postdoctoral Research Fellow at the U.S. Geological Survey investigating the integration of official and crowdsourced geographic information during disasters.
©2012 ACM 1072-5220/12/0500 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2012 ACM, Inc.