Dynamical systems framework for modeling language change in populations of learners with arbitrary social network structures.
Python
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
.gitignore
NBRecreate.py
POC.py
README.md
V2_3params.py
applyT.py
calcP.py
calcTfromextensions.py
exp_adoptercats.py
exp_corridor.py
exp_singleg.py
exp_smallpops.py
generateA.py
setup.sh
updateB.py

README.md

NetworksAndLangChange

Dynamical systems framework (cf Niyogi & Berwick 1997, 2009) for modeling language change in populations of learners extended to represent arbitrary social network structures.

Setup

Just run setup.sh or create a plots directory in the parent directory. Nothing to install.

Dependencies

Everything was implemented in Python 2.7. Numpy is the only external library required to run things.

Experiment Files

There is no user interface yet. Files of the form exp_*.py reproduce plots provided in our papers and serve as code examples

  • exp_adoptercats.py Neutral change among Roberts-style adopter categories. Vary the size of each category and the number of backward edges. contra Blythe & Croft 2012

  • exp_corridory.py Modeling the characteristic two-peak spread of NCS in the St. Louis Corridor and testing Friedman 2014's hypothesis

  • exp_singleg.py Multi-clustered network structure yields S-curve change in populations of single-grammared speakers cf Yang 2009 cot-caught merger

  • exp_smallpops.py Demonstrating the combined effect of small networks and single-grammar speakers on neutral (cf Kauhanen 2016) and advantaged change

Model Files

These implement the model

  • applyT.py Apply T learner transition matrix

  • calcP.py Calculate P learner input matrix matrix

  • calcTfromextensions.py Automatically create T matrix from grammar extensions

  • generateA.py Generate A adjacency matrix

  • updateB.py Update B grammar expression matrix

Other Files

  • POC.py Model proof of concept implemented in python/numpy

  • NBRecreate.py Reproduces some results from Niyogi & Berwich 1997

  • V2_3param.py Inputs for calcTfromextension.py** to reproduce Gibson & Wexler 1994