人类往往通过视觉渠道来了解世界的海量信息，如万物的样貌、形状、颜色、状态，视觉是人类极为重要的感知方式。人类文明诞生之后，能够用于记录、交流、传承的符号（文字）形式同样是“可看”的，视觉符号极大促进人类文明的发展。这也侧面说明，视觉不只是单纯的生理、物理活动，还与人类的抽象认知功能有直接的关联。这种联系投射到语言系统中则表现为：每种语言都具有表达视觉行为的词汇，它们不仅可以描述动作行为本身，而且能够用于其他多种领域，在语言表达实践中扮演着重要角色。

      本文从词汇类型学的理论框架出发，针对6种语系30种语言的视觉动词样本分别进行定名学、符意学研究。定名学的研究思路是“由意义到形式”，核心任务是视觉语义内容被哪些语言形式打包。主要从Talmy的“宏事件”视角和“虚拟位移”理论出发解析了视觉行为过程，在不同认知因素框架下考察每种语言的成词情况，最终得出结论：视觉核心动词存在“一分”“二分”的成词体系，而子类视觉动词的成词等级为（由大到小）：方式类视觉动词 > 隐蔽类视觉动词 > 空间定位类视觉动词 > 速度类视觉动词。

符意学的研究思路是“由形式到意义”。以某意义的视觉动词为基点，考察不同语言中动词负载的多功能语义的异同。本文采用语义地图模型，以“看”“见”为例，展现了每种语言视觉动词的多义扩展情况，发现每种语言在行为相关域、心理认知域、行为目的域、特征相关域、语义虚化域的覆盖节点范围各有差异，整体语系可能存在一定的倾向性特征，如非汉藏语系的“看”明显向行为相关域扩展，“见”向行为目的域的扩展倾向更加显著。且这些语义扩展路径能够被大量共时、历时语料事实验证，显示了人类认知规律共性。最后结合已有成果分析可知，推动语义跨域发展的机制可能有：隐喻机制、转喻机制、主观化机制及重新分析、类推等语法化机制。

Human beings often use visual channels to understand vast amounts of information about the world, such as the appearance, shape, colour and state of everything, and vision is an extremely important form of human perception. After the birth of human civilisation, the symbols (characters) that could be used to record, communicate and transmit were also 'visible', and the creation of visual symbols contributed significantly to the development of human civilisation. It is also clear that the act of seeing is not simply a physical activity, but is also directly linked to the abstract cognitive functions of human beings. This link is reflected in the language system: every language has a vocabulary for expressing visual behaviour, which can be used not only to describe the action itself, but also in a variety of other domains, playing an important role in the practice of linguistic expression.

In this paper, the theoretical framework of lexical typology is used to study the onomasiology and semasiology of visual verbs in a sample of 30 languages from six language families. The idea behind the study of onomasiology is "from meaning to form". The core task is which linguistic forms the visual semantic content is packaged. The main focus of the study is on Talmy's “Macro-Event” and “Fictive Motion” theory, which analyses the process of visual behaviour and examines the lexical formation of each language in the framework of different cognitive factors. The final conclusion is that there is a"one-word" and "two-words" system for the core visual verbs, and that the word formation hierarchy for the sub-categories of visual verbs is (from largest to smallest): Modal visual verbs > Covert visual verbs > Spatial visual verbs > Velocity visual verbs.

The idea behind the study of semantics is 'from form to meaning'. Using a particular type of visual verb as a base, we examine the similarities and differences in the multifunctional semantic load of verbs in different languages. This paper uses a semantic map model to show the multi-sense expansion of visual verbs in each language, using "look" and "see" as examples, and finds that each language has different coverage nodes in the behaviour-related, psycho-cognitive, behavioural purpose, feature-related and semantic deficit domains. For example, the non-Hanzang language "see" has a clear tendency to expand the behaviour-related domain, and "meet" has a significant tendency to expand the behavioural purpose domain. These semantic expansion paths can be verified by a large number of co-temporal and ephemeral corpus facts, showing the commonality of human cognitive laws. Combined with the results available, it is clear that the mechanisms driving semantic cross-domain development are, among others, metaphorical mechanisms, metonymic mechanisms, subjectivation mechanisms and grammaticalisation mechanisms such as reanalysis and analogy.