Executive control in analogical reasoning: Beyond interference resolution

摘要: Executive Control in Analogical Reasoning: Beyond Interference Resolution Anna Chuderska (anna.chuderska@uj.edu.pl) Adam Chuderski (achuders@apple.phils.uj.edu.pl) Institute of Psychology, Jagiellonian University al. Mickiewicza 3, 31-120 Krakow, Poland Abstract In recent studies, analogy-making has been shown to depend on the ability resist interference. our large-sample corre- lational study we found that efficiency analogical reasoning is strongly related measures diverse executive control functions, far beyond interference resolution. These functions included resolution itself, but also: goal activa- tion and application, inhibition response tendency, controlled visual search. The results implicate an important factor for (and probably most types relational) accounting inter-individual differences may be assessing psychological plausibility computational models analogy-making. Introduction Analogy-making comprise high-level cognitive processes structuring two phenomena (analogs, e.g. objects, events, ideas) by similar relations, even if they totally differ seman- tically or perceptually, inferring new goal-relevant information about one phenomenon (target) from elements second (source) means finding out systematic relational correspondences between them. Ability reason analogy essential face any novelty needs understand (e.g., discovery) produce problem solving). Since seminal structure- mapping theory (Gentner, 1983) proce- sses have received much attention science contributed more general theories reasoning. major postulate these processing representations highly dependent limited working memory (WM) capacity. Relational complexity quantifies this limit with relations – number interacting variables must processed parallel. simul- taneously, higher task’s difficulty less people can solve it (Halford, Baker, McCredden, & Bain, 2005). However, are active WM, bigger challenge arises integrate them according a goal, especially conflict across structural semantic constraints (e.g. Markman Gentner, 1993). So, seems probable very capacious WM not enough successfully analogize. For example, LISA artificial neural network model (Hummel Holyoak, 2003), which dynamically binds roles (i.e., arguments) fillers arguments’ content) into synchrony firing their distributed representations, contains intrinsic capa- city (phase set size), since only finite such role-filler bindings oscillate asynchronously cycle. Thus, units enter current phase monitor its goal-progress. This realized lateral competing overall setup influencing way activation spreads. If discriminatory weak, would efficient than counter- part same phase-set size, strong inhibition. indispensible need chose input manage distraction make some researchers Viskontas, Morisson, Hummel, Knowlton, 2004) believe fundamental source human rela- tional comes effectiveness over during reasoning, instead sole storage role begins explored detail. Moreover, despite fact representing manipulating repeatedly differentiating participants (into those who able right analogy), there relatively little curiosity exploiting identifying mechanisms underlying Following others, that, averaging individual group data, leads loosing information, careful examination variability reaso- ning linking other tangible lower-level abilities promising path discerning cogniti- ve (Lewandowsky Heit, 2006). (EC), also referred as control, defined mental states directly, influence organi- ze context internal goal. Execu- tive involved novel situations, when arbitrary sequences responses emitted, great amount planning required, errors likely quickly corrected, dominant relevant tendencies overcome. Recent assume EC emergent process arising dynamical interaction many independent, elementary Braver, Gray, Burgess, 2007). A popular approach identify functions. taxonomy proposed Miyake, Friedman, Emerson, Witzki, Howerter (2000) includes three functions: updating domi- nant responses, shifting. Updating consists conti- nuous “refreshment” contents inserting stimuli while deleting others. Inhibition deals volitional stopping task- irrelevant thought tendency. Mental shifting involves frequent alternating task-sets WM. Other often