主题：An Axiomatic Theory of Situation Attention Utility：From Perception to Choice
主讲人：Chew Soo Hong 西南财经大学
主持人：郑筱婷 暨南大学
时间：2026年9月23日（周三）下午16:00-18:00
地点：暨南大学石牌校区经济学院大楼（中惠楼）305室
摘要
Every choice arises in a situation. Situation Attention Utility (SAU) brings choice-relevant perception into the primitives of decision theory through a situated outcome (x, η): a consequence x, together with a descriptor η recording sensory and contextual features of how it is encountered. The theory studies a certainty equivalent C(X) on finite strings of situated outcomes interpreted as equiprobable lotteries. Under Consistency with Certainty and solvability, two pooling axioms Betweenness and Half-Mix Consistency-characterize C by the ratio-linear form：
where the utility function v is continuous, strictly increasing, and common across situations,while the strictly positive attention function A may depend jointly on consequence and
descriptor.In SAU, situations may reweight consequences without revaluing them. Within a regular reusable architecture retaining each block's certainty equivalent and one positive additive
statistic, Betweenness of pooling and Consistency with Certainty characterize SAU without Half-Mix Consistency. The coordinate change S = Mv(C)yields an additive cache (S, M): a continuation-minimal summary preserving exactly the information required for current and subsequent evaluations. Pooling X and Y combines their caches without reopening their constituent entries:
Applications address correlation-sensitive and context-dependent choice and distinguish attentional loss aversion from behavioral loss sensitivity. Reusable block aggregation also supports the composition of concurrent situations under divided attention and the incorporation of newly discovered outcomes as awareness expands. Experimental evidence is classified by the level of situation sensitivity it can discriminate, yielding testable implications at each level. In recurrent environments, encounter and perception are costly, while the discovery of new possibilities and the complexity of situations call for risk attitudes that vary with the situation. SAU meets both demands economically-non-uniform attention concentrates processing on action-relevant information, cache reuse avoids re-peated evaluation, and a newly discovered outcome is incorporated by appending to the cache and may be favored over more processing-intensive implementations of expected utility when these fitness benefits exceed potential decision losses.
主讲人简介
周恕弘（Chew Soo Hong），新加坡籍华人经济学家，现任西南财经大学智慧经济科学研究中心主任，新加坡国立大学经济系荣休教授。他是世界计量经济学会（Econometric Society）、经济学理论促进学会（SAET）和经济科学学会（ESA）fellow，被誉为行为经济学和实验经济学领域的领军人物。周恕弘的研究领域横跨决策理论、行为经济学、实验经济学、生物经济学及神经经济学。他是公理化决策理论的先驱之一，尤其在非期望效用理论方面做出了开创性贡献。近年来，他致力于探索经济学与生物学、神经科学的交叉融合，提出了“智慧进化论”（Intelligence Darwinism）等新理论，解释人类决策行为的生物机制。他曾获颁四川省人民政府“天府友谊奖”，以表彰其在国际学术交流与合作方面的杰出贡献。
校对 | 郑筱婷
责编 | 彭毅
初审 | 李仲达
终审发布 | 何凌云
（来源：暨南大学经济学院微信公众号）
