Sure machines are also human-invented tools, but their biggest difference from previous tools is that they are what humans must always consider how to communicate with. The most distinctive feature of machines is automaticity: cogwheels engage with one another and hence a chain of action. You only need to provide the motive power first and not to bother with the operation afterwards. That is quite different from previous tools. Even the most proficient cook with the best chef’s knife won’t possibly just sit by and let it chop vegetables itself, I guess?

The transition from mechanical machines to electronic machines rendered a problem how humans should provide the motive power. The way machines receive command wasn’t through mechanical force any more, but through information represented by electric currents. Of course humans don’t discharge electricity (extracorporeally), so it was now a question how to convert human signals into currents.

The earliest way humans communicate with electronic machines was through punch cards. There are only two states: punched and un-punched. And there’re probes in the machine. Slot one punch card into the machine, and they’ll pass through the holes and close the circuits. Of course it’s unelectrified where it’s unpunched, though. So that’s how the earliest human-manchine interaction happened: different circuit combinations meaning different machine action.

But a big problem arose then — the all-too-delayed feedback. You must think through the whole process because these machines will perform your commands all at once. Humans are after all humans, and they just cannot think that thoroughly. For, say, an extremely complicated idea, they have to test it out and modify it, especially when the machine has given its feedback.

So the DOS command line input was thus born. Keyboard input in fact. You know, the codes programmers write and the Matrix-style green characters rain on the black screen are just examples of this. It naturally saved us much trouble. You don’t have to write out the whole program on the punch cards in one go but could see the feedback on the screen while programming. And you could correct your errors immediately you find them. No longer will you run to the machine, pull out the punch card, check hole by hole and slot it back in.

A striking feature of command line input is its linear space. Just think how the cursor moves when you type: it can only do so left-to-right and top-to-bottom, or right-to-left and bottom-to-top. It just cannot jump all the way to the characters several lines away.

So if you have typed for quite a while and find an error above, then to correct it you definitely have to position the cursor where you’ve got it wrong. You need to push Backspace. Unfortunately you have to keep pressing it if the error is far away from where you are. Despite the hotkeys for line skipping, the cursor itself can only move Back and Forward. You fret, looking at the error, but the cursor is still on the way.

Wouldn’t it be fantastic if you can somehow select your errors instantly? And that’s how the computer mouse appeared. Although monitors did exist in the DOS era, they have been totally different since. The monitor for command line input adopts an ant’s perspective, while the GUI kind for mouse inputs a peng’s (鹏) [2], or a bird’s eye view. Only after the invention of the mouse did we first feel the synchronization of our perception with the machine. Think how you tilt your body when playing shooters, which would have been absolutely impossible in the command-line-input era. When you grip the mouse, your body won’t treat the game scene before you as from another world, even though you are perfectly sensible.

[2] The peng (鹏) is a legendary enormous bird in Chinese mythology, much like the ziz and the roc.

The computer mouse has therefore enabled a new freedom. However, it still follows some trajectory, whether moving straightly or crookedly. Would it be possible if the cursor could just “blink” elsewhere? That’s actually what the touch screen is, the major input device for smart devices for now. The transition from “trajectory” to “blink” is also one of human operation from click to touch.
With the transition of our gesture comes also the transition of its ambience. In my childhood playing on the computer was actually an occasion. It should be indoors, where you’ve got power supply. And it should be at a stretch, so preferably at weekends, sometimes even when you’ve finished your homework. In conclusion, the excessive formality created a sharp distinction between virtuality and reality.

The evolution from the punch card, through the command line and the computer mouse, to the touch screen, is indeed the gradual expansion of the screen from the 0-D, through 1-D and 2-D, to 3-D world[3]. It’s also the gradual erosion of the distinction between reality and viruality. For the biggest dream of the screen is to absorb the whole real world.

[3] The touchscreen is in essence the mapping of 3-D gestures to the 2-D plane.