Investigating within-speaker variation in the QuakeBox Canterbury earthquake stories


UC CEISMIC is proud to announce that a paper utilising data collected by the UC QuakeBox project will be presented at the 43rd New Ways of Analyzing Variation conference (NWAV43) to be held on October 23-26 in downtown Chicago, Illinois. The paper will be presented by Lynn Clark, Jen Hay, and Liam Walsh, of UC's New Zealand Institute of Language, Brain and Behaviour, along with UC CEISMIC's Paul Millar. The conference is hosted jointly by the University of Illinois at Urbana-Champaign and the University of Illinois at Chicago.

Their abstract reads,

In the aftermath of the Canterbury earthquakes of 2010-2011, everyone who had experienced the quakes had stories to tell.   People's experiences were diverse, and often dramatic.  In the months following the quakes, people would tell their 'Earthquake story' often.  We wanted to create a collection of these stories.  Our motivations were threefold.   First, many people felt strongly that they wanted their stories to become a part of the public record, and be available for subsequent generations to learn from.   We felt that capturing, transcribing, and making available these stories would be an important community service.  Second, a collection of recorded earthquake stories would provide a valuable research content repository for researchers across many disciplines, interested in investigating the manifold personal and societal impacts of the earthquakes. Third, a research archive containing many 'danger of death'-like monologues , all describing the same time and event, would be of particular value for sociolinguistic analysis.

The UC QuakeBox project was therefore formed, as part of a collaborative project between the New Zealand Institute of Language, Brain and Behaviour and the UC CEISMIC group. The QuakeBox was a shipping container, which was converted for use as a transportable recording studio.  We positioned the QuakeBox at various locations in and around the city of Christchurch, and invited the public to record stories of their experiences of the 2010-2011 Canterbury earthquakes.  By the end of 2012 the QuakeBox project had recorded 722 earthquake stories.

These recordings have been carefully transcribed and time-aligned in ELAN (at the utterance level), and then force-aligned with HTK (at the phoneme level).  The corpus in its entirety contains approximately 120 hours of recordings, with many participants treating the QuakeBox as an opportunity to speak candidly. Consistent with our impression that many people wanted to share their stories publicly, a total of 576 of the 722 stories were flagged by participants for full video release on the publicly-accessible UC CEISMIC Canterbury Earthquake Digital Archive website.  The collection is a remarkable set of highly engaged and engaging stories, delivered by speakers who enthusiastically volunteered to contribute their story to the archive.

In this paper, we introduce and demonstrate the QuakeBox corpus, and outline some of the rewards and challenges associated with collecting stories in a manner that was purposefully and saliently in the public eye.  We also present the results of the first linguistic analysis conducted on the corpus.    This is a careful variationist study of medial /t/, the realization of which varies considerably in NZ English.    We focus particularly on patterns of phonological repetition or priming within speakers,  a topic which has received little attention in variationist sociolinguistics but which the QuakeBox corpus is particularly well suited to addressing, as it contains a series of unusually long and  engaged monologues.    This presents the opportunity to investigate patterns of variation within individual speakers in a way that is not affected by interaction patterns with interlocutors, nor by rapidly changing topics.

It's exciting to see the UC CEISMIC archive already being used to enable new research, only a few years into its existence.  We're also pleased to see that our archive is of value not just to earthquake researchers, but can be used by a wide range of disciplines to research questions we never imagined when CEISMIC was first conceived.  The UC CEISMIC team are looking forward to seeing what other research emerges in future years.