本文主要討論《易》學詮釋中的象數問題。象數和義理是兩種截然不同的詮《易》方法。「象數」詮《易》有許多重要的預設，最基本的當然是預設宇宙是一個規律性的、機械性的結構，然後再運用象數理論以解釋此一結構中的規律。至於義理詮《易》則不外兩種模式－即王弼模式和程頤模式。這兩種模式都旨在尋求象數體系無法思辨的超越層面、屬於本體論層次的義理。唯前者處於東漢末，以漢儒為論敵，不自覺反受漢儒之影響，故人文意義尚未深發；後者則以儒家人文思想為本，遂為近世義理《易》典範之作。本文從「天人關係」的觀點，提出「天衡定人」和「人衡定天」兩種視界，區分了象數詮《易》與義理詮《易》的不同。其次提出十種「象」和三種「數」，並說明象數理論是原本《易經》編撰時即已存在的思想，不是漢儒的新創。象數詮《易》更涉及古代歷史、天文學、曆法學等不同範疇的知識與問題，在古代文化史中，自有其不可磨滅的效用與價值。總之，義理派對於象數理論的批評，並未能徹底推翻象數之說，但象數理論所包含的許多不符合自然科學理論的部分，研究者也應該客觀地加以檢討、揚棄。

Theories of interpretation of the Yijing (Book of Changes) can be summarized as two different approaches: the philosophical approach and the "xiangshu" (images and numbers) approach. This paper describes the differences between these two approaches, and, concentrating on the latter method, further defines ten categories of images and three categories of numbers of the Yijing. Through analyzing and comparing the examples within the commentaries of the Han philosophers, Wang Bi and the Song philosophers, this paper draws the following conclusion: Rather than being invented by the Han Confucians, the xiangshu theory came into existence when the Yijing was edited. The hexagram and line texts comprised different kinds of mystical numbers and images that were too effective and valuable, for ancient traditions of historians, astronomers, and calendricists, to be neglected in cultural history. The xiangshu approaches and the philosophical approaches are two parallel lines with no conjunction since both of them are based on some specific concept of world order. Wang Bi criticized the xiangshu theory of the Han philosophers, but he also limited himself unconsciously by their xiangshu theory so that his cosmological concern overcame his concern for humanistic values.