tsimane-glottal-interspeech23

〈’〉 in Tsimane’: a Preliminary Investigation

This code was used to generate the results reported in the InterSpeech2023 paper entitled "〈’〉 in Tsimane’: a Preliminary Investigation"

Code

Production

When a command generates files, they will be saved in the same directory as the script generating them. The filename of the generated files will show the Git SHA hash (12 characters) of the data set and the time and date at which the file was generated.

From the root of the repository, run the following command which will read the TextGrid files and extract metadata. This command generates the file that is used for the statistical analyses.

python -m scripts.interspeech2023.batch1_glottal_duration

Then open analyse_glottal_duration.Rmd, modify hash_date so that its value matches that of the generated file and run the analyses.

Perception

collect_results.PraatScript will collect the results of the perception experiment located in tsimane-glottal-qui.

analyses.Rmd will generate the result table presented in the paper.

Figures

To generate the figures (spectrograms), go to the figures directory and execute the following command

bash make_figures.sh

Note that this command will only work if Praat is installed on your computer.

Cite

To cite our work use the following BibTex (temporary until published)

@inproceedings{havard-et-al-glottal,
  year = {2023},
  month = aug,
  publisher = {{ISCA}},
  author = {William N. Havard and Yaya Sy and $ Camila Scaff and Loann Peurey and Alejandrina Cristia},
  title = {{〈’〉 in Tsimane’: a Preliminary Investigation}},
  booktitle = {Interspeech 2023}
}