IBSS: New Methods for Investigating the Formation of Individual and Shared Concepts and Their Dynamic Dispersion Across Related Societies

这一跨学科研究项目将开发和测试新的数学模型，探索随着语言复杂性的变化，概念意义在语言中的表达方式。项目调查人员还将检查社会中的个人群体之间分享这种含义的方式。该项目的模型将把概念的动态发展描述为一个几何系统，并将提供方法来理解概念的语言表示，以及一个社区的语义如何受到邻近社区的影响。虽然这个项目将集中于颜色术语在语言和社会中的演变方式，但这个项目的洞察力和信息将超越颜色表示的领域，应用于任何一组概念，其中对象具有可以用数学方法评估和描述的相似结构。将在本项目期间开发的办法和方法将在以下情况下发挥作用的例子如下：(1)开发明确和正式可扩展的人工智能和机器人类比人类分类和分类系统；(2)开发可用于增强快速全球信息消息传递能力的全球通信方法；(3)为医疗诊断系统和运输系统等关键系统中的信息表示建立标准化系统。概念的形成和交流渗透到各种人类活动中。它们在教育系统、运输系统的组织和设计、零售市场和消费品的实体和虚拟设计、医疗诊断的质量和风险分类、企业业绩和社会价值方面发挥作用。心理学家、语言学家、人类学家、计算机科学家和其他学者通过关注概念形成的具体例子来研究概念，同时试图理解这样的概念系统是如何形成的。一个受到关注的具体概念是颜色术语是如何“演变”的，以及它们的概念含义如何被社会成员理解和分享。颜色概念化是一个重要的特例，因为颜色刺激可以精确测量和容易复制，并且可以用数学精度来描述人类百万种颜色的感知空间。这个项目将专注于数学模型的开发和测试，以捕捉颜色术语概念的分类和共享方式。要设计和测试的模型将使用几何形式来描述一般意义，并将通过研究颜色术语和概念来具体演示它们的用法。测试将使用来自各种社会的数据，包括中美洲颜色调查(MCS)，这是一个系统收集900个人的分类行为的数据库，这些人曾与处于颜色词典发展不同阶段的116种濒危或发展中语言中的一种或多种进行交流。MCS主要是手写形式的，该项目的一个产品将是利用现代计算机科学的众包方法实现其完全数字化。监测和控制系统数据库的完全数字化将使其首次可供全球科学界使用。该项目通过NSF跨学科行为和社会科学研究(IBSS)竞赛得到支持。

This interdisciplinary research project will develop and test new mathematical models that explore the ways through which conceptual meaning is represented in languages as those languages change in complexity over time. The project investigators also will examine the ways such meaning is shared among groups of individuals in societies. The project's models will describe the dynamic development of concepts as a geometric system and will provide methods for understanding the linguistic representation of concepts and the ways semantic meaning from one community can be influenced by that of neighboring communities. Although this project will focus on the ways that color terms have evolved within languages and societies, the insights and information from this project will apply beyond the domain of color representation to any set of concepts in which objects have a similarity structure that can be assessed and described mathematically. Examples of the kinds of situations where the approach and methods to be developed during this project will have utility are the following: (1) the development of unambiguous and formally scalable artificial intelligence and robotic analogs of human classification and categorization systems; (2) the development of a global communication methodology that could be used to enhance rapid global information messaging capabilities; and (3) the construction of standardized systems for information representation in critical systems, such as medical diagnostic systems and transportation systems. The formation and communication of concepts permeate a diverse range of human activities. They play roles in education systems; in the organization and design of transportation systems; in the physical and virtual design of retail markets and consumer goods; in classifications of quality and risk in medical diagnoses; in business performance; and in social values. Psychologists, linguists, anthropologists, computer scientists, and other scholars have studied concepts by focusing on specific examples of concept formation while trying to understand how such conceptual systems are formed. One specific concept that has received attention is how color terms "evolve" and how their conceptual meaning is understood and shared by members of a society. Conceptualization of color is an important special case because color stimuli can be precisely measured and easily duplicated, and the human perceptual space of a million colors can be described with mathematical precision. This project will focus on the development and testing of mathematical models that capture the ways color term concepts are categorized and shared. The models to be designed and tested will use geometric formalisms for characterizing meaning in general and will specifically demonstrate their use by investigating color terms and concepts. Testing will use data from a wide variety of societies, including the Mesoamerican Color Survey (MCS), a database of systematically collected categorization behaviors of 900 individuals who have communicated with one or more of 116 endangered or developing languages that are at various stages of color lexicon development. The MCS is largely in hand-written form, and one product of this project will be its full digitalization using modern computer science crowd-sourcing methods. Full digitalization of the MCS database will make it available for use by the global scientific community for the first time. This project is supported through the NSF Interdisciplinary Behavioral and Social Sciences Research (IBSS) competition.