This project began with a desire for privacy while writing in public, and quickly became a much bigger undertaking that reshaped how I approach design, taught me how to make fonts, and created an opportunity for a pretty unique kind of glyph-based artwork. It's still a work in progress, but I want to present what I've done so far because this has been some of the most fun I've had with an art project in a while.

How it started

Over the summer, I spent hours every week on the train to work. After a few days of staring at my phone for two hour-long trips per workday, I started bringing books and a notebook along so I could rest my eyes, practice my handwriting, and get rid of some distractions. This was a great change for me and I began to look forward to those train rides as my writing and reading time. However, there's one downside to writing in public: people can read over your shoulder very easily.

I wasn't writing much that I actually needed to keep hidden, but the lack of privacy got me thinking about something I've had in the back of my mind for a long time: what if I could invent a writing system that would act as my own personal privacy filter? This wouldn't be particularly difficult, I'd essentially just be creating a handwritten font with symbols that could be substituted for existing English letters and numbers. But, as is my custom, I decided to take this challenge as seriously as possible and develop a system of symbols that made logical sense and adhered to consistent rules.

Designing the alphabet

I arrived to most of these rules by observing patterns that arose in my initial doodles. As you can see in the image above, I started by drawing semi-random shapes until one idea began to stick. I'll come back to how I arrived at these rules in a moment, but I think it's best to explain what I landed on first.

Here are the rules:

1. Every number and letter must only consist of combinatations of 13 predetermined lines, similar to a digital clock's 7-segment display
2. No letter can use all 6 of the outermost lines (that when drawn alone would create a rectangle). Similarly, every number must use all 6 of those lines
3. Every letter must have at least one closed area, none of which can be rectangular
4. Punctuation cannot have any closed areas

Those became the rules that dictate the consistency of the font, but there are also guidelines that aren't as set in stone. The main one is to try to evoke the English letter equivilent as close as possible without sacrificing the design language. The idea is to make the text readable to anyone who puts a little time into familiarizing themselves with it while remaining nearly unreadable to the uninitiated. I also wanted a focus on symmetry because I thought the system would be prettier if many of the symbols had sister characters that were reflections of themselves.

I won't explain every choice I made in this process, but here are a few example of how I approached resolving conflicts between similar characters. The current symbol for N ( N ) was originally A ( A ), but I realized at some point that A should use the other diagonal so that N could have the diagonal that's already present in its English character. B ( B ) and K ( K ) were the same characters with a small difference for a while, but after fully committing to the theme of symmetry I decided to make them similar shapes but flip K horizontally because it's rarer. I ( I ) and T ( T ), and O ( O ) and Q ( Q ) were some of the characters I figured out relatively early, but at the last minute I decided to switch the two sets with each other after reading I as O and vice versa too many times.

Designing the number system

The numbers also deserve a moment of recognition because I'm very proud of this design. I wanted to hide the shape of each Arabic numeral in their symbol while also creating a unique system to remember the unavoidably abstract characters. In the system I settled on, 0 is naturally the simplest ( 0 ), and each set of 3 afterwards creates a set. The lines from 1 ( 1 ) and 2 ( 2 ), when overlaid, make the character for 3 ( 3 ), 4 and 5 make 6 ( 4, 5, 6 ), and 7 and 8 make 9 ( 7, 8, 9 ).

Creating a digital font

I developed this font nearly to its current form by hand, but towards the end of the summer I began to use my computer on the train more, and I quickly realized how nice it would be to use the system I was already growing familiar with to read and write on my computer as well. I've been interested in making fonts for a long time, and with years of experience in Adobe Illustrator, some basic knowledge of kerning principles (more on that soon), and the free and open source font design software FontForge already installed on my computer from a few halfhearted attempts in the past, I started my journey.

Making the shapes in Illustrator was time consuming but ultimately easy. All I needed was to make each of the 13 possible segments, then attach them to each other in different ways to create each character I had already designed. Once I had each character exported, I brought them into FontForge to compile the font. I quickly ran into issues that uncovered a world of font creation madness that I had no prior awareness of, but after a whole lot of troubleshooting and learning and re-exporting every character way too many times, I was finally able to work on the best part: making the font look pretty.

A tangent about kerning

Kerning is the custom spacing between pairs of characters that make them fit together more naturally than they would with even spacing (called monospacing in the world of typography). I had already figured out most of the pairs of letters that fit together, albeit with an early version of the font. A good example of the importance of kerning are the letters "PL" ( PL ) versus "PO" ( PO ). "PL" has the font's standard spacing, but because the "P" has that overhang on the right side, the "O" fits neatly under it if the spacing is less than the default. This font has comparably few possible outlines compared to most, so I had a unique opportunity to go all out with some pretty extreme kerning because so many of the glyphs fit together like puzzle pieces.

Pleasing words and finding art in limitations

Once I had an incomplete but usable digital font, I began to dive into the artistic potential of this project. There are many letter combinations that fit together perfectly (NV ( NV ), POJ, ( NV ), and HXE ( HXE ) are just a few examples), and I spent a while one afternoon trying to find words that were particularly pleasing to look at. CIVILITY (CIVILITY), BLOODWOOD (BLOODWOOD), SOCIOBIOLOGICAL (SOCIOBIOLOGICAL), and COMMITTED (COMMITTED) are some of my favoites. The theme of pleasing words, by the way, seems to be some combination of those well-fitting letter groups I've taken to calling puzzle sets, adjacent sister glyphs or double letters, and highly symmetrical words. It was also fun to discover that there's absolutely no correlation between words that are typically thought of as beautiful, and the ones that look visually pleasing when simplified by this system. RENNAISSANCE (RENNAISSANCE) is quite beautiful but EPIPHANY (EPIPHANY) is rather messy.

When individual words have such potential to be visually pleasing, the natural next step is to create artwork that isn't bound by the rules of the English language. My favorite use of these glyphs in artwork so far happens to be the simplest: lines of repeated puzzle sets that resemble various stitching patterns. It evokes a surprisingly handmade feel for something so clearly digital, an effect I'm sure I couldn't have created intentionally. I've tried laying out the letters in various different ways in 2D space, but I can't seem to find anything as pleasing as those simple stitching lines. Maybe one day I'll embroider a bag or something with little glyphs that translate to complete gibberish.

Something that was reaffirmed for me throughout this project is my long-held belief that the character of all kinds of art and technology is partially created by their limitations. In the years I spent most of my time in film, countless project-defining creative choices were made because of the lack of resources at our disposal. This includes everything from playing multiple characters in a middle school short film due to having a small crew, to choosing to follow fewer subjects in a documentary due to, well, having a small crew, and so much more. Similarly, the things that make people nostaglic for old technology are the things that have been engineered out of their modern replacements, such as the scratchiness of vinyl records or the tinny sounds of AM radio. This project has been defined in a few ways by removing limitations on my resources, most notably in the fact that I learned how to use an a professional font creation software instead of finding a more restrictive but less time-consuming way to make a font. However, early on in this project, I realized that by creating such strict rules for the design language, I was opening the door to a unique and geometric artwork that wouldn't have been possible if the letters were as thematically disconnected as they were in my first drawings. So although my resources have grown throughout the process, in choosing to keep the design language strict for reasons of both pragmatism and personal taste, I artificially created an environment in which I could find the character (pun absolutely intended) of this project by using those self-imposed limitations to my advantage.

What's next?

I'm still working on making the complete suite of puncuation and other symbols, but once that's done, I may make the font available on this page for anyone to download. No promises though. I also think there's a lot more potential for cool artwork with the patterns that appear throughout this font than I've discovered to date, and I have a hunch that this will become even more of a visual art project than I ever expected.

I also might change the name, because "Angular Glyphic" was meant to be a placeholder but I got kind of attached to it.