Abstract
I present a computational-level model of semantic interference effects in online word production within a rate–distortion framework. I consider a bounded-rational agent trying to produce words. The agent's action policy is determined by maximizing accuracy in production subject to computational constraints. These computational constraints are formalized using mutual information. I show that semantic similarity-based interference among words falls out naturally from this setup, and I present a series of simulations showing that the model captures some of the key empirical patterns observed in Stroop and Picture–Word Interference paradigms, including comparisons to human data from previous experiments.
Highlights
In cognitive science and related fields, bounded rationality is the idea that our cognitive systems are designed to take actions that are approximately optimal, given that only limited computational resources are available for calculating the optimal action (Simon, 1955, 1972; Kahneman, 2003; Howes et al, 2009; Lewis et al, 2014; Gershman et al, 2015; Lieder and Griffiths, 2019)
The main contribution of this paper is to show that rate–distortion theory generically predicts the well-documented semantic interference effects that a subject experiences when trying to produce a target word in the presence of a semantically related distractor
The Stroop task famously exhibits interference (Stroop, 1935): given a stimulus, such as the word BLUE printed in red ink, and an instruction to name the color of the ink, it is hard to produce “red” because of interference from the similar word “blue.” A similar kind of interference is present in the Picture–Word Interference task, where a drawing must be named in the presence of a superimposed distractor word (Lupker, 1979; Starreveld and La Heij, 2017)
Summary
In cognitive science and related fields, bounded rationality is the idea that our cognitive systems are designed to take actions that are approximately optimal, given that only limited computational resources are available for calculating the optimal action (Simon, 1955, 1972; Kahneman, 2003; Howes et al, 2009; Lewis et al, 2014; Gershman et al, 2015; Lieder and Griffiths, 2019). A bounded-rational action policy is a policy that chooses an action to maximize some measure of reward, or equivalently, to minimize the cost of the consequences of taking a certain action in the world, subject to a constraint on the computational resources used in finding and implementing this action. These resources include factors, such as time—in many circumstances, it may be more important to act quickly than to take the time to compute the best action—as well as physiological resources, such as the energy required to perform computations. Letting D(s, a) represent the action cost or the cost of the consequences of taking action a in state s, and letting C(s, a) denote the computation cost required to compute the action a given state s, the overall cost for a policy q can be written as
Sign-in/Register to view highlights & summary 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 callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
