Bayesian pragmatic models in Python

Models

• The basic Rational Speech Acts model of Frank and Goodman 2012
• The lexical uncertainty model of Bergen et al. 2012
• The anxiety/uncertainty model of Smith et al. 2013
• The anxious experts model of Levy and Potts 2015
• The streaming lexical uncertainty model useful for large problems like those in Potts et al. 2015

To see these models at work on an example involving the division of pragmatic labor, run

python -m pypragmods.pragmods

which runs the main method example given in full at the bottom of the file. In essence, if one has created a set of lexica lexica, and used it to instantiate a Pragmod called mod, then the different models are accessible with

mod.run_base_model(lexica[0])
mod.run_uncertainty_model()
mod.run_anxiety_model()
mod.run_expertise_model()

The current version is compatible with Python 2 and Python 3.

The disjunction code

For examples of the anxious experts model in action, see disjunction/bls41.py.

python -m pypragmods.disjunction.bls41

It includes the code for the illustrative examples in Levy and Potts 2015 and Potts and Levy (reference below). In particular, the function compositional_disjunction shows how to use lexica.py to create a space of lexica for analysis with pragmod.py.

Embedded scalars code

The code in embeddedscalars implements the compositional lexical uncertainty model of Potts et al. 2015. The core pragmatic models is in pragmods.py; this code creates a logical grammar (fragment.py), implements functions for refining that grammar (grammar.py), analyzes our experimental data (experiment.py), and reproduces all of the figures and tables in the paper (paper.py, making use of analysis.py for the comparisons between model and experiment. For examples, paper.py is the best place to start:

python -m pypragmods.embeddedscalars.paper

The subdirectory experiment contains the experimental code and materials. For additional guidance on how to use these materials, see the repository for Dan Lassiter's Submiterator.

References

Bergen, Leon; Noah D. Goodman; and Roger Levy. 2012. That's what she (could have) said: how alternative utterances affect language use. In Naomi Miyake, David Peebles, and Richard P. Cooper, eds., Proceedings of the 34th Annual Conference of the Cognitive Science Society, 120–125. Austin, TX: Cognitive Science Society.

Frank, Michael C. and Noah D. Goodman. 2012. Predicting pragmatic reasoning in language games. Science 336(6084): 998.

Levy, Roger and Christopher Potts. 2015. Negotiating lexical uncertainty and expertise with disjunction. Poster presented at the 89th Meeting of the Linguistic Society of America, Portland, OR, January 8–11.

Potts, Christopher; Daniel Lassiter; Roger Levy; Michael C. Frank. 2015. Embedded implicatures as pragmatic inferences under compositional lexical uncertainty. Ms., Stanford and UCSD.

Potts, Christopher and Roger Levy. 2015. Negotiating lexical uncertainty and speaker expertise with disjunction. To appear in Proceedings of the 41st Annual Meeting of the Berkeley Linguistics Society.

Smith, Nathaniel J.; Noah D. Goodman; and Michael C. Frank. 2013. Learning and using language via recursive pragmatic reasoning about other agents. In Advances in Neural Information Processing Systems 26, 3039–3047.