Code for Bayesian model of the co-evolution of language and mindreading

• In the folder python_code in this repository you can find all the Python code I used to implement a Bayesian model of the co-evolution of language and perspective-taking/mindreading (including development, iterated learning and biological evolution) (see Woensdregt et al., 2016 for a paper on the developmental part of this model). I subdivided the code into the core modules (files without a prefix), the main scripts for running a full simulation (files prefixed with run_), and the scripts for analysis and plotting (files prefixed with unpickle_).

• In the folder code_for_running_on_cluster you can find the code you need to run a batch of simulations as an array job on a Grid Engine computer cluster (works on the Open Grid Scheduler batch system on Scientific Linux 7 at least). The example allows you to create an array job on a cluster which runs the same simulation (2 independent runs of a single learner learning from a single speaker for 30 observations) but looping through each of the possible speaker lexicons (for a lexicon with 3 meanings and 3 signals in this case; 343 possible lexicons in total).

References

Woensdregt, M., Kirby, S., Cummins, C. & Smith, K. (2016). Modelling the co-development of word learning and perspective-taking. Proceedings of 38th Annual Meeting of the Cognitive Science Society.