10 Years After Katrina, the Enduring Value of the Hurricane Digital Memory Bank

(Originally posted on the RRCHNM blog, August 28, 2015)

Ten years ago, we knew as historians that we couldn’t assess fully the social, cultural, economic, and political implications of the devastating hurricanes in the summer of 2005. We did know that previous natural disasters had profound consequences. The 1927 Mississippi River Flood, for example, further fueled African American migration to northern industrial cities, and paved the way for federal intervention in southern states during the New Deal. Documenting the reactions and memories of individuals affected by Katrina, and then Rita, along the Gulf Coast, took on an urgency soon after the storms hit.

Michael Mizell-Nelson, the late-public historian from the University of New Orleans, reached out to CHNM’s late-director Roy Rosenzweig to discuss the possibilities of creating a community-sourced digital project to document the aftermath and recovery of Hurricane Katrina. With so many residents relocating, collecting online gave anyone who had been displaced an opportunity to share their reflections and document their stories. This became even more important following Hurricane Rita three weeks later, when some Gulf Coast residents evacuated a second time, some never returning home.

The Center collaborated with the University of New Orleans to form a team that built and promoted the Hurricane Digital Memory Bank <hurricanearchive.org> (HDMB) in the fall of 2005. HDMB played an important role in documenting the stories of individuals, and offered a non-commercial digital space to collect photographs, audio diaries, or digital video that would be cared for as a digital collection. The team hoped that the process of telling one’s personal story could help in the healing process of individuals, families, and communities, while also expanding the voices documented and accessible in the historical record.

When we began collecting, we wanted HDMB to complement, not replicate other documentary efforts.

HDMB was designed in the tradition of the September 11 Digital Archive, and intentionally different than a traditional oral history project. An online collecting project offered some important advantages. First, we collected all of our materials in a single digital platform accessible to anyone with an Internet connection. All submissions were vetted and available within 48-36 hours. Second, we reached a wider audience by permitting contributions from anyone and at any time—this was especially important in documenting a storm that has dispersed Gulf residents across the United States. Third, contributors decided what they wanted to share and in what format: a personal reflection in the form of a poem, digital photographs, or emails. Fourth, an online project allowed us to reach more people, more economically. Finally, the materials collected were in a digital form, making this content data available for future computational analysis.

Some scholars remained skeptical of our open-door collecting policies that allowed anyone to contribute to a historical collection. Skeptics wished the project’s creators controlled the quality of the submissions by carefully selecting participants and the voices saved and accessible.

When carefully planned, a community-based approach to collecting can allow for a more diverse sampling. Projects like HMDB or the September 11 Digital Archive offer a unique raw archive of sources. The team did not alter or correct metadata received with items. Online contributions are representative of the individuals who submitted them. In the collections section, we asked those submitting large numbers of files to describe what they gave us–at the collection or item level.

The project’s ability to make sources available quickly online for researchers caught the attention of individuals who turned to the project to care for their memories and collections. As a result, we became the place where an individual could share her story as she moved across the country, like the late-Courtney Giarrusso, and where historians who lacked resources at their home institutions shared their own photographs, video diaries, and oral histories, such as the curators at the National Hansen’s Disease Museum. Institutions that invested in large-scale oral history projects also deposited copies of their collections with HDMB, such as the University of Southern Mississippi’s Center for Oral History and Cultural Heritage, to increase their use. We received over 900 photographs taken by staff at the Smithsonian National Museum of American History during a collecting trip that haven’t been accessible through Smithsonian online collections. The volunteer organization, Katrina’s Kids Project, brought art supplies to children living in the Reliant Astrodome shelter in Houston, Texas, and wanted the art work saved for the historical record. These items now only live inside HDMB. As a result of our open-door policy, HDMB holds unique collections. RRCHNM remains committed to preserving these individual contributions and collections, keeping them accessible online. A preservation copy of all items in HDMB will be deposited in George Mason University’s MARS digital repository.

Digging Deeper

As mentioned above, one of the advantages of collecting online first, is that we have content ready for computational analysis. Researchers can always discover online holdings by browsing or searching through the collections, but also through analyzing the digital content at scale. Searching through collections of primary source may identify a few key sources worthy of a closer reading, while text analysis of the entire site can offer a researcher a greater sense of the entire holdings and the themes represented within.

Easy-to-use text analysis programs, such as the Voyant Tools help researchers cull through the bulk of the collections by surfacing word patterns and trends that might trigger further investigation. It is possible to see word frequencies and view relationships of terms in context with others. Not surprisingly, place names featured prominently in individual contributions. To look beyond the names of cities, parishes, or states, it is possible to create a list of “stop words” that removes those terms from the analysis. Without place names, it is possible to see that HDMB’s contributors frequently mentioned “people,” “home,” “house,” and “family.” By examining those keywords in context, it is possible to see how mentions of “house” relate to the descriptions of physical damage and destruction. While usages of “home” often discuss the emotions of leaving or returning to a damaged house or city. It is possible to identify other emotional terms, such as “loss” and “angry,” and see that “hope” is invoked as a verb and a noun more often than both loss and angry. In thinking through what other patterns might become visible, topic modeling can be applied to the same texts.

It is possible to then run that corpus through a light-weight topic modeling tool. All item descriptions can be extracted and analyzed or it is possible to focus on the writing within the online stories. The text can be run through the topic modeling tool multiple times to refine the stop word list, while also seeing how the software makes connections across this corpus. It is possible to see that this digital collection would be useful for someone interested in reading first-hand accounts about evacuations, life in shelters, volunteerism, the challenges of returning home to deal with damages, the emotional challenges faced in the recovery process, financial burdens faced by storm survivors, and the impact of local, state, and federal government in a disaster. With such topic strings identified, the tool then allows the researcher to drill down and read the individual texts that were grouped together.

Researchers do not need to read all 1650 online story contributions to find the larger themes. They can carefully dip into these collections, from the topics discovered, and pull back as more questions surface. We welcome this type of analysis of the content inside HDMB.

Legacies

Since HDMB launched, it has been used by scholars, students, and journalists as a valuable research collection. Scholars writing about the short-term political, social, and cultural effects of Katrina and Rita published articles in Southern Spaces, the Journal of American History, and other edited collections with HDMB in their bibliographies. Doctoral students interested in learning about disaster recovery and relief efforts turned to HDMB to read individual accounts. The format and structure of HDMB and other digital-first collections have been written about by archivists and oral historians. This month, the Odgen Museum of Art has asked their visitors to contribute HDMB as part of their temporary exhibition about remembering Katrina. We hope that researchers will continue to turn to HDMB, particularly as we move farther away from these events and begin to gain a wider historical perspective.

This week as the nation remembers the impact of Hurricane Katrina and its aftermath on communities and individuals, and on the larger national memory of the American people, RRCHNM also remembers two of HDMB‘s creators, Roy Rosenzweig and Michael Mizell-Nelson, as well as collaborator David Shayt, who are no longer with us to remember and reflect.

If you have wish to reflect on the anniversary, the collecting portal remains open.

Ten years later, HDMB honors all of their legacies by remaining a freely accessible unique digital collection of historical sources will be sustained for historians and researchers in the years to come.