Code and Data Accompanying the Study "A New Framework for the Representation and Computation of Uncertainty in Phonological Reconstruction"

This repository provides code and data accompanying our study on the representation and computation of uncertainty in phonological reconstruction.

List, J.-M.; Hill, N. W.; Blum, F.; and Forkel, R. (forthcoming): Representing and Computing Uncertainty in Phonological Reconstruction. To appear in: Proceedings of the 4th Workshop on Computational Approaches to Historical Language Change.

To get started, install all required packages:

$ pip install -r requirements.txt

Having installed all packages, we recommend to first download the data. This can be done with the Makefile we provide:

$ make download

For the individual commands (you will need to have git installed), please refer to the Makefile.

To run the analyses for each language, you can again use the Makefile:

$ make karen
$ make burmish
$ make panoan

This will run all analyses and specifically create Markdown and PDF files that show the output of the analyses. All results are also provided here.

The statistics reported in our study can be computed with the following command:

$ make stats

The result we obtain is:

Dataset	Prediction	Count	Proportion	Alignment Size
Karen	correct	246	0.65	4.03
Karen	false	133	0.35	4.27
Karen	certain	310	0.82	4.05
Karen	uncertain	69	0.18	4.41
Burmish	correct	154	0.57	4.13
Burmish	false	115	0.43	4.29
Burmish	certain	199	0.74	4.13
Burmish	uncertain	70	0.26	4.39
Panoan	correct	405	0.79	4.25
Panoan	false	109	0.21	5.14
Panoan	certain	465	0.9	4.37
Panoan	uncertain	49	0.1	5.14