log in
Cursive Scripts are the Road to Awe, Part 1
2021-03-08 08:36:37

Cursive Scripts are the Road to Awe, Part 1

There is a unique delight to be had in looking at lines of inscrutable, confident cursive prose in a language you cannot begin to grasp. Cursive scripts present an allure of cohesion and sophistication, asserting that the writer is comfortable with the language and is going through the motions of using it to communicate ideas; conversely, block lettering suggests the written text itself is a challenge or accomplishment to put forth. So it goes with the evolution of real scripts, and so it goes with our fantasy languages.

The transition—from etched and scribbled runes to coiled and snaking paragraphs—presents some new obstacles to refined and consistent production, and makes urgent certain deficiencies that went hitherto unnoticed. So, in this very special follow-up to Your Writing System Sucks, we shall discuss the nuts and bolts of how to keep your conlanging pen flowing with the wind.


Consider the following lines, drawn with a mouse:

The human arm and hand cannot produce perfectly straight, linear movement without a tremendous amount of very specific practice. Some people master this art after considerable effort—and gain infinite kudos with their uncanny ability to perform the "robot" dance—but it is not a talent normally expected of writers. When I drew the above lines with a mouse, I did so without the use of my wrist, and with the base of my hand stabilized on a flat surface. The arrows show which why my hand was moving at the time. Some of the lines are quite straight—mostly because they were drawn slowly, which is highly undesirable in cursive writing. Initially it doesn't look like there's any sort of systematic bias between leftward and rightward strokes, but piling them up produces a somewhat clearer pattern:

In both directions, my elbow exhibited a tendency to push out too far at the start of the stroke, then retract slightly too far afterward. This is a result of the reciprocal action of the muscles in my arm, which operate as a pair of class 3 levers:

When writing with a pen, you also have the benefit of additional muscle control from your wrist, but most likely your arm will be less stable, even if you pin your elbow against your torso. As a result, if you repeat the above exercises with a pen on paper, you probably won't see the same pattern; instead, you'll get four different levers and two different rotors interacting to approximate a straight line as best they can.

Obviously, I don't have the muscle memory of a technical illustrator! If I drew letters as wobbly as I drew these lines, my writing would look like a toddler's. (It doesn't.)

There are still clearly some fairly strong biases at play, and when writing cursive we are usually working with short, non-linear motions that do not have these problems.
In several of these lines, we seen an "M" or "W" shaped wobble, where one phase of the elbow joint's movement is cancelled out by a phase of the wrist joint's movement. This was also evident in some of the mouse examples.

When writing extended lines of text in a fluent cursive script, the baseline will show a tendency to wobble according to the "elbow-only" curve unless you make effort to reposition your hand during writing.

If this sounds like evidence of the inferiority and frailty of human flesh, rest assured that machines using a basis of curved motions such as these are vastly superior for energy efficiency and reliability, and that every engineer since Archimedes (and probably far earlier) has accepted this.


1. No stroke is ever going to be perfectly straight. In cursive scripts, we exploit this as part of the design.
2. It is essential to understand the tools being used in order to produce something coherent and attractive.
3. The more joints are involved in making a hand motion, the more attenuated (controlled) the line gets.
4. When only using the elbow to draw a horizontal line, a sigmoid (S-shaped) bias is evident. For biomechanistic reasons, it has the same general shape as the sine function, but we can control its amplitude (scale) and phase (horizontal offset) dynamically, mainly through practice and slow movement.
5. When also using the wrist joint, we're adding a second rotor, which also generates a trigonometric function. In a horizontal line, we want this to cancel out the curvature caused by the elbow joint. The resultant line will be shaped as the sum of these two functions, which is the source of the "W" or "M" shape.
6. The shoulder, logically enough, adds another trigonometric function to the mess.

Straight lines in other directions are still subject to these processes. Curved lines take advantage of the underlying biomechanical phenomena rather than fighting them, and consequently have more consistent and smooth results. But, as we'll see in the next part, they're never perfect, and we must use more techniques to produce good-looking final results.
Samantics comment   8456.94 tgc / 2021.182 ce