Links' traits make them interesting signs of the author's "presence" within the structure of the hypertext document. For one, links exist within nodes and yet between them. In their capacity within a node, a link signifies a point of departure, lack, and rupture. Between nodes, they serve as a means of navigation, association, and contact. In each sense, a link functions within contexts which are always under the control of the author (see Space 3). That is, the author decides to provide a possible rupture in her text at any given point, and she seizes the opportunity to associate two nodes for any "reason" she likes and at any point. Another aspect of links can be considered in Charles Peirce's language of "icon," "index," and "symbol." Links may function in any one or in several of these three dimensions. And how they are represented within these dimensions depends upon the author's conception of the previously mentioned relationships between nodes. Each of these traits can be used by the web author to facilitate or subvert the reader's experience through the hypertext.

Beyond that, I think that in at least one sense, Peirce's conception of signs makes room for an "ethical" aspect in the way signs function. Obviously, links within a hypertext are "signs" in the common sense of the term sign, as a unit within a language system which denotes meaning through differentiation within that system and through reference to some "other" or signified--that is, links possess the power of signification--they signify at least the meaning "you can leave this node by clicking here." Signs tend to be seen as relational functions connecting a signified to a signifier in some various ways, depending upon which structuralist you ask. For Peirce, the sign's "essential function" is to "render inefficient relations efficient--not to set them into action, but to establish a habit or general rule whereby they will act on occasion." Links do this just as I described above--they establish a general rule whereby they will provide the efficient change of nodes on the occasion that the link is clicked. Peirce claims "Knowledge in some way renders [relations] efficient; and a sign is something by knowing which we know something more" (643-4). Without stretching it too thinly, this comment applies especially to links as a special, hypertextual case of signs. By knowing a link--knowing its presence in the document, how to use it, and its context--one can arrive at more material to know. A link within a hypertext document signifies lack, an incomplete "knowledge" within the existing node. It suggests a relation to another node which will allow the reader the option of "knowing more"; how it renders this relation is its ethical dimension.