Abstract
How do people perceive and communicate structure? We investigate this question by letting participants play a communication game, where one player describes a pattern, and another player redraws it based on the description alone. We use this paradigm to compare two models of pattern description, one compositional (complex structures built out of simpler ones) and one noncompositional. We find that compositional patterns are communicated more effectively than noncompositional patterns, that a compositional model of pattern description predicts which patterns are harder to describe, and that this model can be used to evaluate participants’ drawings, producing humanlike quality ratings. Our results suggest that natural language can tap into a compositionally structured pattern description language.
Highlights
Humans see patterns everywhere, and eagerly communicate them to one another
We assessed whether participants in the drawing part of the study used more dots to redraw compositional than noncompositional patterns
These results suggest that there is an interface between natural language and the compositional pattern description language uncovered by our earlier work (Schulz, Tenenbaum, et al, 2017)
Summary
This article seeks to bridge this gap by combining a pattern communication game with a mathematical model of pattern description (Quiroga, Schulz, Speekenbrink, & Harvey, 2018; Schulz, Tenenbaum, Duvenaud, Speekenbrink, & Gershman, 2017). Consider the graphs shown, which plot time series of CO2 emission, airline passenger volume, and search frequency for the term “gym membership.”. Experiments suggest that humans perceive these graphs as compositions of simpler patterns, such as lines, oscillations, and smoothly changing curves (Quiroga et al, 2018; Schulz, Tenenbaum, et al, 2017). There is seasonal variation in passenger volume (a periodic component with time-dependent amplitude), superimposed on a linear increase over time
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
