Program Execution As Theatre Play

What does it mean to say that the execution of a program is like a manifestation of a play? The first thing I think about is that the user is part of the play, but indirectly. Peter Bøgh Andersen writes:

A computer-based play consists of signs, each sign characterized by attributes and a sequence of actions, most of which the audience can hear or see but some of which it must infer.¹

So the signs are the actors in the play, but the user isn’t an actor. Instead, the user is seen as influencing the signs. This is done through the “handling features” of a sign. The other two classes of attributes are “permanent features” and “transient features.”

A handling feature, Andersen writes, “is produced by the user and includes key-press, mouse and joystick movements.” So some signs are interactive and can be controlled or affected by the user. This could be signified by a transient feature, for instance that a button is drawn differently when it is pushed, or that a checkbox is checked when an option is chosen. A transient feature might also be the result of being influenced by other signs. A permanent feature, on the other hand, “is a property of the sign that remains constant throughout [its] lifetime,” something that contrasts the sign against others, and thus identifies it.

The curtain falls and signs appear. The user (in the audience) affects an interactive sign, which in turn influences other signs. The states of the signs change, which is reflected by transient features. The user perceives this and again influences her interactive sign. And so on.

This reminds me of a computer game called The Game of Harmony, where the interactive sign is a sphere which can move and bounce other spheres into each other. When same-colored spheres collide they disappear. If they are of different color, new spheres are created. On higher levels, spheres are joined by “rubberbands,” so that they pull others along as they move. Then there are obstacles in the form of walls to complicate things.

In his typology of computer-based signs, Andersen lists the classes “actor sign” and “object sign.” I’m unsure how he would classify the other spheres in Harmony. Perhaps they are too passive to be classified as actor signs. Both actor signs and object signs has permanent and transient features, and lack handling features. But actor signs, he writes “still [have] some action associated to them,” whereas object signs “cannot influence other signs, but can themselves be influenced.” But then again, he classifies the ball in Breakout as an actor sign, so he probably would include the spheres in that category.

Other classes in the sign typology are “controller signs,” and “layout signs.” Controller signs would in Harmony be the walls, because they “change other signs,” or bounce off spheres in this case, “although they do not change their own visual appearance.” And layout signs “lack transient and handling features, and have no functions vis-à-vis other signs,” so the “sky” or the backdrop would qualify as such a sign.

One interesting thing in Harmony is that when a sphere moves past the edge of the screen, it “wraps around” to the other edge. If it is joined with other spheres with rubberbands, the rubberband wraps as well, and the tension is often increased manifold, causing now distant spheres to come shooting quickly towards the first sphere. If the other spheres are of different color, many new spheres will be created.

The question is whether the edge would qualify as a controller sign or as a “ghost sign.” A ghost sign is one that “[lacks] both permanent and transient features [but] do have function to other signs,” in this case to teleport the sphere to the opposite edge. A controller sign would have a visual representation, but as the edges of the screen can’t be drawn, does that disqualify the sign as being classified as a controller? If the game would be run in a window instead of in fullscreen mode, would that make a difference?

¹ Peter Bøgh Andersen, A Theory of Computer Semiotics.