504 Voices New Orleans English Corpus

Description of Project

The 504 Voices Project was a data collection effort enacted by Katie Carmichael (Virginia Tech) and Nathalie Dajko (Tulane) starting in 2018 thanks to support from the National Science Foundation (BCS-1749217). You can access grant information here. The 504 Voices Project has a separate website where we provide general information about the project alongside information about New Orleans English. It is also where we host the web version of our Dialect Awareness Curriculum for New Orleans teachers; you can access the PDF version here. The 504 Voices Corpus consists of data collected for that project, along with additional interviews collected by Katie Carmichael in 2016-2018, as well as Carmichael’s dissertation interviews collected in Chalmette in 2012. Information about the corpus is provided below. At the bottom of the page is a google form to request access.

Why 504 Voices?

504 (pronounced five-oh-four) is the area code for New Orleans, and is often used as shorthand for the city. This corpus provides examples of voices from the 504.

• Date Collected: 2012; 2016-2023

• Language: English

• Location: New Orleans, Louisiana (Orleans Parish) and Chalmette, Louisiana (St. Bernard Parish)

• Speaker Information: Total of 192 speakers; full demographic information available here

  • New Orleans subcorpus: 135 Black, white, and Creole speakers. 64 men and 71 women. Birthyears: 1917-2002. All grew up in New Orleans, Louisiana. Collected in 2016-2023 by Katie Carmichael, Nathalie Dajko (Tulane), Dana Serditova (Uni Regensberg), Lucia Paternostro (Tulane), Shawanda Marie (Independent).

  • Chalmatian subcorpus: 57 white, working-class speakers. 32 women and 25 men. Birthyears: 1927-1994. All grew up in and around the town of Chalmette in St. Bernard Parish, though half of the corpus had relocated after Katrina (most to St. Tammany Parish on the Northshore of Lake Pontchartrain, some elsewhere in Greater New Orleans). Collected in 2012 by Katie Carmichael.

• Recording Information: Participants wore Shure SM 10A headset microphones and were recorded by a ZOOMH4 portable digital recorder, a separate Crown Audio Sound Grabber II microphone was set up to record other speech and interaction. All force-aligned files isolate the headset microphone audio, though interview recordings where the interviewer is audible are also available upon request (these would not be transcribed/force-aligned however).

• Tasks: Most participants completed an interview, reading passage, and word list. Thus for most participants, files available include: headset mic-isolated WAV audio file, TXT manual transcription, and PRAAT textgrid with time-aligned phonetic transcription available for 3 speech conditions.

  • Full interviews consisting of casual conversation lasted one to four hours; 15-45 minute segments selected and transcribed as TXT file and FAVE-aligned. 

    • At the end of the interview, metalinguistic commentary was elicited; some of this portion of interviews is transcribed in MS Word and may be made available upon request. 

  • Reading passage and word list data available for most participants, both of which were also FAVE-aligned. 

Description of Files

Interview Speech

• File types: txt, force-aligned textgrid, wav

• Each participant completed an unstructured interview consisting of an oral history about their lives and views on New Orleans. 

• In total, there are over 84 hours of recordings with an average transcribed interview speech length of 26 minutes and 20 seconds.

• Manually transcribed in txt, force-aligned with MFA into PRAAT textgrids (hand-corrected)

Reading Passage

• File types: txt, force-aligned textgrid, wav

• Each participant was instructed to read the following reading passage aloud.

• Link to Reading Passage (2022 version)

• Link to Reading Passage (original version)

Word List

• File types: txt, force-aligned textgrid, wav

• Participants were instructed to read this word list aloud multiple times through.

• Link to Word List (2022 version)

• Link to Word List (older version)

I request that if you publish using this data set, you cite/acknowledge its use and make me aware of the publication via e-mail. I also ask that if you further develop the data (e.g. transcribing or force-aligning parts of interviews that are not yet processed) that you share that with me so I can make it available to other researchers and continue building on the data set.

How to Cite

Carmichael, Katie & Nathalie Dajko. (nd). 504 Voices New Orleans English Corpus.  https://sites.google.com/vt.edu/vtlx/corpora/504-voices-new-orleans-english-corpus  

Contact Information

katcarm@vt.edu

Corpus Request Form

Access the form here.