Word-initial consonant lengthening is a robust phenomenon across typologically diverse languages

If you use some of the code, please cite the following article:

Blum, Frederic and Paschen, Ludger and Forkel, Robert and Fuchs, Susanne and Seifart, Frank. Accepted. Consonant lengthening marks the beginning of words across a diverse sample of languages. Nature Human Behaviour.

Preparing DoReCo for the analysis

This workflow leads you through the process of reproducing our analysis. We use certain system dependencies that we assume to be installed, like sqlite3, git, Python, and R. We recommend to run all the code in a fresh virtual environment. The code was tested using cmdstan version 2.32.2 and might not work with other versions. Please make sure to have the correct version installed.

To start, please first clone this repository to your workspace and cd into the directory.

git clone https://github.com/Tarotis/initial_lengthening
cd initial_lengthening

The first step consists in cloning the repositories of the DoReCo data, as well as CLTS. You can clone them with the following commands:

git clone https://github.com/cldf-datasets/doreco
git clone https://github.com/cldf-clts/clts --branch v2.2.0

The next step is to download the DoReCo and convert it to CLDF. To do this, first cd into the DoReCo repository and install all the necessary packages.

cd doreco/
pip install -e .

You can now run the download and the conversion to CLDF. While downloading, insert "Y" when asked for ND-data. You do not need to download the audio files, and leaving them out will speed up the downloading. When prompted for the path to CLTS, type ../clts/ to link to the cloned CLTS-repository.

cldfbench download cldfbench_doreco.py
cldfbench makecldf cldfbench_doreco.py --glottolog-version v5.0

The last step created a SQLite database out of the CLDF data, to quickly access all the data. Once you have the SQLite database ready, you need to install the pre-written views. Run SQL:

cldf createdb cldf/Generic-metadata.json doreco.sqlite
sqlite3 -echo doreco.sqlite < etc/views.sql
cd ../  # On Windows, you might have to use the full path to the preceding folder

You can now run the query we provide to extract all the data needed for the analysis. Please note that we make use of the stdev function which is only available if you run the query through doreco.query, and not through other standard interfaces.

cldfbench doreco.query --format tsv init_query.sql > scripts/data.tsv

Running the analysis

Please cd into the scripts-directory:

cd scripts/

Now you are in a folder full of R-Scripts. The first script 00_setup.R includes a groundhog-call, which installs packages with the correct versions. Running this script will create a folder R_groundhog in the current working directory which includes those packages. It also includes calls to the other scripts. You can run all of them via the following command:

Rscript 00_setup.R

Please not that the two scripts which run the brms-model are currently commented out. This has a simple reason: Running them takes several days. If you want to re-run the model, please go to the respective script and run it from there, or un-comment the line in 00_setup.R. We provide the fitted model within our OSF-repository here so that you do not need to re-run the model.

If you want to run code from within the individual R-files, please set the working directory to the scripts-folder so that all the code is run correctly.