« Recession Era Victory Garden in Oakland, Pittsburgh | Main | Improvised Explosives in Iraq: Of makers, disarmers, and the dead. »
June 8, 2009
How people find a common language
Humans carrying out a task together find a common language, sooner or later. Even communities of speakers do this, and in fact this how they may come up with natural languages over time. At Carnegie Mellon University, we have now simulated this process in a cognitive computer model, tracing the steps that humans take to learn the elements of a graphical language, to forget them, to re-learn, and to arrive at a common language in a small community. Psychological research provides a pretty precise picture of how human memory works. In this paper to be presented at the International Conference on Cognitive Modeling, we describe the combination of two worlds: multi-agent and evolutionary simulation on the one hand, and psychologically validated models of memory and cognition to simulate the evolution of a domain language. The model explains the empirical data pretty well, but also makes a prediction: that human communication between two partners needs to go both ways in order for us to learn and to converge.
David Reitter and Christian Lebiere.
Towards explaining the evolution of domain languages with cognitive simulation. In: Proceedings of the 9th International Conference on Cognitive Modeling (ICCM), Manchester, UK, 2009.
Abstract
We simulate the evolution of a domain language in small speaker communities. Data from experiments (Garrod et al., 2007; Fay et al., 2008) show that human communicators can evolve graphical languages quickly in a constrained task (Pictionary), and that communities converge towards a common language even in the absence of feedback about the success of each communication. We postulate that simulations of such horizontal evolution have to take into account properties of human memory (cue-based retrieval, learning, decay). We implement a model that can draw abstract concepts through sets of non-abstract, related concepts, and recognize such drawings. The knowledge base is a network with association strengths randomly sampled from a natural distribution found in a text corpus; it is a mixture of knowledge shared between agents and individual knowledge. In three experiments, we show that the agent communities converge, but that initial convergence is stronger when communities are structured so that the same pairs of agents interact throughout. Convergence is weaker in communities when agents do not swap roles (between recognizing and drawing), predicting the necessity of bi-directional communication in domain language evolution. Average and ultimate recognition performance depends on how much of the knowledge agents share initially.
Posted by dr at June 8, 2009 6:44 PM