Click to see Map. If you click and it doesn't appear, check your taskbar (top or bottom of screen usually) to see if that window is already open.

Links considered in Charles Peirce's language of "icon," "index," and "symbol."
Peirce divides signs into three categories, "icon," "index," and "symbol." I want to appropriate these terms to establish a means of distinguishing the functions links can serve within a hypertext document. Knowing the possible functions for links and having a specific terminology for those functions helps when considering the ethical options that an author has when writing a hypertext.

Iconic function: Graphic images can be made "hot" upon the screen. However, to remain within Peirce's terms, not all images are truly icons. In fact, only those images which represent through their resemblance to the material of the linked node. Further, the icon function of a single link does not necessarily require the link be presented by an image to be iconic. According to Greenlee, Peirce accounts for icon signification as a resemblance which serves "to help direct attention to other properties of the object than those the sign-vehicle is conventionally interpreted to represent" (79). To show this, Peirce suggests that two algebraic equations which consist of similar characters--"especially when we put resembling letters for corresponding coefficients"--resemble one another as an icon. I contend that as a single link, that is, as a single unit which functions in a specific way--beyond any symbolic content within the link--the iconic function of a link can be invoked by the link's resembling in words the relationship manifested in the connected node. So both a picture of a cat and a single link stating, "cats eat mice" are functioning iconically as links if the node connected to each clearly resembles the qualities exemplified in the link.

Indexical function: In Peirce's system, "index" identifies a sign which shows an acting upon the indicator by the object it indicates. As examples, he offers relatively clear cases such as symptoms of diseases, weathervanes, and barometers; however, Greenlee notes that Peirce's continual defining of this term has turned it into a rather confusing concept. For purposes of discussing an "indexical function" of links, I will work with the seemingly more basic and accepted interpretations of index. From this standpoint, links possess three basic indexical functions. First, in the sense of the weathervane, which points as the wind actively directs it, a link can be programmed to react to the browser's pointer passing over it. I am not referring to the finger that appears, that is an indexical function of the browser, and cannot be accounted for in the programming of the webpage itself. Instead, I refer to the status bar change I have been using all along--or the ability to do this--or even something a little You touched it!

Second, the link can indicate the clicking of the mouse in two ways: it changes colors, and it calls a new node onto the screen. Finally, a web designer can program the links to indicate whether the user has already seen the node they call onto the screen. This indexical quality is the most questionable to me. When one visits a node which is connected in a many to one relationship, all of the links connecting to that node will exhibit change. Greenlee, while not discussing "index" directly, offers an example which seems to allow this third function: if a hunter sees a tiger track in the jungle, and interprets them

"as signs of a tiger in his surroundings. The sign is significant only on the condition that the interpreter has learned by previous experience of the connection between tigers and their tracks. . . . [but imagine that] the marks in the ground taken for tiger tracks are in fact blurred impressions, the cause of which is not a tiger. The hunter is mistaken: the object he takes the impressions to be a sign of is fictive . . . built out of past experiences of tigers. What this almost amounts to is that the immediate object IS the past experience as relevant to the interpretation of the present sign. . . . The point of reducing the immediate object to 'relevant previous experience' is to obtain a conception of the object of the sign which is illustrated by any instance of signification." (68)

In this quote, Greenlee is engaged in a much larger argument than I need to include, but the useful result of this comment for understanding the indexical function of a link is that the reader can be mislead by an author's not including contextual information that can counterbalance the potential for mistaken "past experience."

Symbolic function: Peirce's final category of signs, the symbolic, is possibly the most common type of link mechanism. At one level, the symbolic function of a link is typically the simple word or phrase that suggests some association with another node. However, an author can use the symbolic function by establishing a color code and marking links within that code. In fact, that links are differentiated, by default, by colors and by underlining (when they are simple typographical characters) is an example of the symbolic function enacted by browsers. However, unlike the case of a browser's index function that shows a pointing finger when it encounters a link, I argue that the default color function is indicative of the author's desire for or inability to change the standard colors. The underlining feature seems to be a symbolic function of the browser and not of the author's control.

Of course image based links can have symbolic functions as well. The basis for symbolic signs in Peirce relies upon their meaning being generated by convention. Clearly words and numbers com quickly to mind, as language is a matter of convention. However, as I noted above, there seems to be a sense in which, in links, words can be used as iconic representations of larger, but closely similar nodes. Likewise, though images are typically used iconically in hypertext, they can quite easily be seen as used in symbolic functions. Greenlee suggests the use of the "Mona Lisa" as a symbol for all art.

For my purposes, I try to work out some of the more obvious and useful conceptions of the functions on a separate window table.