Many paradigms in cognitive science posit that human learning is characterized by a limited capacity to represent the information relevant for a given task. We argue that excess capacity -- using more representational resources than needed for a task at hand -- is a plausible alternative paradigm for the study of human learning. Leveraging recent results from machine learning, we show that excess capacity can be consistent with high predictive ability. We also review extant empirical findings from the cognitive science literature, demonstrating that excess capacity learning can account for a range of empirical phenomena, such as humans' simultaneous yet apparently contradictory tendency to both memorize observations and capture higher-level patterns in them. We conclude by discussing promising directions for future inquiry under the excess capacity learning paradigm.

We present a novel framework for assessing the fit of cogni-tive models. Using this framework, we highlight limitationsof existing methods of model evaluation, and derive new ap-proaches to validating cognitive models. Tests of topographi-cal consistency emphasize how a model’s structure constrainsbehavior on pairs of coupled stimuli, even when point predic-tions on individual stimuli depend on estimates of the model’sfree parameters. By carefully selecting these coupled stim-uli such that they follow the distinct topography of the model,researchers can overcome some limitations of existing meth-ods. Finally, we provide a proof-of-concept example of how touse our approach to assess a model of multi-alternative, multi-attribute choice.

Framing issues as matters of non-negotiable values can increase the perceived intractability of debates. Focusing on theconcrete consequences of policies instead can facilitate conflict resolution. Using a topic model of Reddit commentsfrom January 2006 to September 2017, we show that the contribution of certain topics concerned with protected val-ues to the debate increased prior to the emergence of a public consensus in support of same-sex marriage and declinedafterwards. These topics related to religious arguments and freedom of opinion. In contrast, discussion of certain con-crete consequences (the impact of politicians stances and policy implications) showed the opposite pattern, their increasedprominence coinciding with improved public support for same-sex marriage after 2012. Our results reinforce the mean-ingfulness of protected values and consequentialism as relevant dimensions for describing public discourse and highlightthe usefulness of unsupervised machine learning methods in tackling questions about social attitude change.

Effective problem solving requires both exploration and ex-ploitation. We analyze data from a group problem-solving taskto gain insight into how people use information from past expe-riences and from others to achieve explore-exploit trade-offs incomplex environments. The behavior we observe is consistentwith the use of simple, reinforcement-based heuristics. Partic-ipants increase exploration immediately after experiencing alow payoff, and decrease exploration immediately after expe-riencing a high or improved payoff. We suggest that whetheran outcome is perceived as “high” or “low” is a dynamic func-tion of the outcome information available to participants. Thedegree to which the distribution of observed information re-flects the true range of possible outcomes plays an importantrole in determining whether or not this heuristic is adaptive ina given environment.

We explore structural issues with parameter estimation fornon-linear cognitive models: Some parameter values are eas-ier to recover than others, and the recoverability of differentparameters interacts in systematic ways. We propose methodsfor researchers to anticipate and visualize and these issues, andthe systematic ways they differ across experimental designs.Our approach consists of assessing how changes in parame-ter values translate into changes in behavioral predictions, anddevelop measurements of the relative responsiveness of predic-tions to parameter values. We demonstrate application of ourapproach to cumulative prospect theory (CPT), a widely-usedmodel of risky decision-making.

We develop a conceptual framework for studying collective adaptation in complex socio-cognitive systems, driven by dynamic interactions of social integration strategies, social environments and problem structures. Going beyond searching for 'intelligent' collectives, we integrate research from different disciplines and outline modelling approaches that can be used to begin answering questions such as why collectives sometimes fail to reach seemingly obvious solutions, how they change their strategies and network structures in response to different problems and how we can anticipate and perhaps change future harmful societal trajectories. We discuss the importance of considering path dependence, lack of optimization and collective myopia to understand the sometimes counterintuitive outcomes of collective adaptation. We call for a transdisciplinary, quantitative and societally useful social science that can help us to understand our rapidly changing and ever more complex societies, avoid collective disasters and reach the full potential of our ability to organize in adaptive collectives.