The people who work on the world’s writing systems are scattered across continents, and we rarely get to sit in the same room. This past March offered a rare exception: two back-to-back European events — a workshop in Liège and a conference in Cambridge — where scholars, type designers, and encoders gathered in person to puzzle through the documentation, design, and encoding of ancient scripts.

At both events, common threads of speculation surrounding fonts, keyboards, encoding models, and standardization schemes for ancient scripts circulated among attendees, and I was there with my notebook ready.

Liège: Deciphering, depicting, and encoding historic scripts

I landed in Belgium on March 3rd for the first leg of the trip. The Ancient writing systems, typography, and encoding workshop, hosted at the University of Liège, was scheduled to begin on March 5th. Organized by François Desset, expert in the Linear Elamite and Proto-Elamite scripts, the workshop brought together experts and scholars ready to hash out the documentation, typography, and encoding of historic scripts.

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

Arriving a bit early meant that I had one full day to explore the city of Liège (and beat the jetlag from California)

After a brief exploration of the University of Liège campus, the first day of the workshop began with an introduction from Thomas Huot-Marchand, director of the Atelier National de Recherche Typographique (ANRT) and one of SEI’s partners in the Missing Scripts program. Thomas set the stage by describing the intersection of the different fields that work on script encoding and how our different activities overlap. Academia, the tech industry, and designers often work in parallel, but workshops like this are sometimes one of the only places where these three worlds collide with the specific intention of discussing solutions. Thomas concluded with an announcement that the 2026 Unicode Technology Workshop will be held in Nancy, France this October, which aims to offer another site of fostering this shared purpose.

This first day of the workshop was dedicated to hieroglyphic scripts, starting with Egyptian hieroglyphs. Stéphane Polis walked through a fascinating history of the different typographic representations of Egyptian hieroglyphs since the early days of their decipherment. Early transcriptions of Egyptian hieroglyphs were very distorted by the cultural influence of the scholars studying the script. For instance, transcriptions of bird glyphs were sometimes drawn in the style of European birds rather than birds native to the region. This level of influence in these transcriptions, which can be thought of as the early “fonts” for Egyptian hieroglyphs, caused a huge variation in the representation of the signs — the repercussions of which can still be felt in today’s fonts for Egyptian hieroglyphs.

With a freshly-illuminated picture of the murky world of Egyptian hieroglyph fonts thanks to Stéphane, Andrew Glass took over the talk to describe a brief history of the script in the context of Unicode. Egyptian hieroglyphs were encoded in phases as technologies emerged and improved, with additions spanning Unicode versions 5.2, 12.0, and 16.0. The result is a slew of characters and controls that rely on a font to do the heavy lifting to “make sense” of the encoding layer and properly render the script.

Much to the awe of the scholars in the audience, Andrew announced that he has successfully created a font for Egyptian hieroglyphs that can render the complex joining behaviors of the script, including quadrats¹ and cartouches². Andrew’s goal for this font, which contains 200,000 lines of code, is for people to use it as the technical foundation to design prettier fonts without having to recreate all of the code needed to execute the combining behaviors.

Andrew and Stéphane demonstrating typing in Andrew’s font for Egyptian hieroglyphs. Users, who are primarily scholars, type combinations of keys that correspond to glyph codes that they are already familiar with in the scholarly community or existing literature. Andrew is seeking feedback from Egyptologists for their thoughts on the keyboard inputs

Thanks to working out the combining behaviors for Egyptian hieroglyphs, Andrew demonstrated that the Rosetta Stone can be fully transcribed into Unicode-compliant text

In the final hours of Day 1, Christian Prager and Alexandre Bassi discussed Maya hieroglyphs. Though occupying opposite sides of the world, Egyptian and Maya hieroglyphs (and hieroglyphic scripts in general) face similar problems in decipherment, transcribing, encoding, and rendering.

One such challenge can be simply cataloguing the sheer number of signs that compose the script. Christian started off the talk by describing his 15-year project to create a database for Classic Maya: a massive corpus containing around 1100 signs with around 1600 graphic variants for Classic Maya hieroglyphs.

Next, Alex took over from Christian to describe his work with SEI to develop a font for Classic Maya. Type designers face unique challenges in creating fonts for historic scripts, which largely depend on the structure of the script and the stage of its decipherment. For instance, because there are so many variants for Classic Maya glyphs, Alex has to converge on the best possible representation for visual glyphs that include strictly the information that matters. This means locating many glyph examples and generalizing those attestations into one representative form. Alex also explained that he does not want the glyphs in his font to refer to a particular time and place in Classic Maya history — a dilemma particularly relevant for ancient scripts like Egyptian and Maya hieroglyphs that span extensive time periods and large geographic regions.

Christian’s text database and dictionary for Classic Maya hieroglyphs (Source: Interdisciplinary Dictionary of Classic Mayan)

Christian walking through some history of Maya hieroglyphs

Example of size variations for a glyph from Alex’s Classic Maya font. Similar to the quadrats in Egyptian hieroglyphs, Andrew Glass developed a 6×6 grid for the font engineering in order to properly render Maya glyph compositions. This means that there are 36 possible variations for each main sign (Photo credit: Alexandre Bassi)

Alex presenting examples from his font for Classic Maya

The day concluded with a group dinner generously provided by the department at the University of Liège. After many laughs and spirited conversations between colleagues who rarely have the chance to socialize in person, it was time to gear up for the final day of the workshop.

Day 2 focused primarily on Proto-Elamite and Linear Elamite, two scripts used in what is now Iran from the Bronze Age to roughly 1850 BC. François Desset, who specializes in the Elamite scripts, began the day with a brief background on Proto-Elamite and an update that he is using his decipherment of Linear Elamite to make progress on deciphering Proto-Elamite. The two scripts have long had a gap in their timeline, but—thanks to recent carbon dating technology—François claims that some inscriptions are from this “middle period”, suggesting a continuity or evolution between Proto-Elamite and Linear Elamite.

For scripts like Proto-Elamite that remain presently undeciphered, font design must accommodate unknown or changing information. Kaveh Ashourinia, who took over the talk from here, explained that he had to create two options for reading direction in his Proto-Elamite font since the directionality of the script is not yet known. Because writing behaviors cannot be observed for ancient scripts, type designers try to replicate writing practices as best as possible. As part of Kaveh’s research, he recreated the writing tools potentially used by Proto-Elamite scribes as a way to gain insight into the stroke direction and letterforms.

François presenting about Proto-Elamite and Linear Elamite

Kaveh presenting about the research and design process for his Proto-Elamite font. This slide demonstrates the different reading direction possibilities for Proto-Elamite. Though the script is undeciphered, Kaveh believes that Proto-Elamite tablets are read top-to-bottom and left-to-right based on the stroke directions of the engraved letterforms

Kaveh’s recreation of the possible writing tools used to write the Proto-Elamite script on clay tablets. Recreating writing practices for ancient scripts can help to inform the encoding model by providing clues about shaping behavior, backspacing, etc (Photo credit: Kaveh Ashourinia)

We then pivoted to Linear Elamite. François provided a background of the script and an overview of the scholarly debates surrounding his proposed decipherment. Under François’ decipherment, Linear Elamite is the oldest known purely phonographic writing system, with 77 phonetic values (revolutionary at the time).

Historic scripts like Linear Elamite have the potential to be volatile; new discoveries of artifacts or innovations in the decipherment can cause changes to the sign inventory or script features, complicating both fonts and the encoding model for inclusion in Unicode. However, many scholars desire Unicode-compliant fonts for such scripts to publish these very discoveries and innovations. This results in somewhat of a “chicken and egg” problem similar to that of newly-invented scripts.³

From here, Sina Fakour presented his font for Linear Elamite, developed using François’ decipherment. Although type designers try to avoid the influence of modern scripts, they also must consider the user community for historic scripts in the design process. Sina described that for ancient scripts, since the use case is mostly academic publishing, you need to design fonts to have visual harmony with whatever script or font is used most commonly by the scholarly user community. Thus, Sina designed his Linear Elamite font to harmonize (in terms of line weights, glyph heights, and spacing) with Times New Roman since Latin and Linear Elamite would appear alongside each other in publications.

Sina’s font “Hatamti” for the Linear Elamite script (Photo credit: Sina Fakour)

Sina presenting his process of superimposing traced examples of a Linear Elamite glyph to converge on the general glyph outline. For historic scripts, it is important to avoid assuming that ancient scribes intended to follow modern design conventions. Properties like symmetry and geometry seem natural to us now, but was it the intention back then? Through this process, Sina concluded that Linear Elamite scribes did indeed intend to have symmetry and geometry, but the key point is that this information must be derived through research, not assumed

François pleading to the Unicode gods

SEI Technical Director Anshuman Pandey wrapped up the discussion by situating Proto- and Linear Elamite in the context of Unicode. Anshu’s presentation really solidified encoding as the “missing step” for these scripts: Sina and Kaveh’s fonts cannot be compatible with Unicode until code points exist for their respective scripts; and scholars like François cannot include searchable, indexable text in their publications until such fonts are Unicode-compliant.

Anshu walked through a number of strategies for encoding historic scripts into Unicode that take their potential volatility into consideration. Since Unicode has a stability guarantee (i.e., once something is encoded, it cannot be removed), one tactic Anshu suggests is to use number codes as the Unicode code point names rather than alphabetical or linguistic names, which may be subject to change as scholarly understanding about the script develops. Information about the sign features (geographical information, sound value) can instead be encoded in the annotations, since those—unlike code point names—can be changed after publication of the Standard. Script proposal authors can consider strategies like these when advocating for the inclusion of ancient scripts in Unicode.

I was just able to squeeze in a group selfie after the talks before I ran off to catch my train, unfortunately missing the final session of the day by Cécile Pierrot. I left Liège with a desire for more opportunities like this workshop to exist — a rare chance for scholars and technical experts to share their work and feedback with each other. Fortunately for me, I was heading to Cambridge for precisely that.

Group photo from the Ancient Writing Systems, Typography, and Encoding workshop at the University of Liège (Photo Credit: François Desset)

Selfie with François, Thomas, Anshu, and Andrew (Photo credit: Thomas Huot-Marchand)

A mutually grateful visit to the lovely city of Liège

Cambridge: User-focused deep dives into scripts around the world

I had the pleasure of spending the last few days of March at the Writing As Visual Engagement (WAVE) conference in Cambridge, UK. Hosted by the Visual Interactions in Early Writing Systems (VIEWS) Project at the University of Cambridge, this conference was the second iteration of the WAVE conference, which previously took place in September 2024. A year and a half later, WAVE 2 was ready to kick off on March 26th for four days of talks and socializing among scholars, type designers, and script enthusiasts from around the world.

This was my first time in Cambridge, as well as my first time meeting many friends of SEI, and I was delighted to immerse myself in the projects of my peers in this beautiful setting.

Since the first day of WAVE 2 began in the afternoon, I had time in the morning to squeeze in some quick views of the city and colleges from the River Cam

Since the first day of WAVE 2 began in the afternoon, I had time in the morning to squeeze in some quick views of the city and colleges from the River Cam

Since the first day of WAVE 2 began in the afternoon, I had time in the morning to squeeze in some quick views of the city and colleges from the River Cam

Trying a Chelsea bun

On my walk to the Faculty of Classics

On my walk to the Faculty of Classics

Day 1 of the conference began with a warm introduction from Pippa Steele, director of the VIEWS Project. Pippa reflected on the first WAVE conference and what a privilege it was to foster that community again this year.

While the first WAVE conference focused on the visual experiences of writing, Pippa explained that the theme of WAVE 2, Writing As Visual Engagement, intends to emphasize the active participation of the user in the practices, processes, traditions, and acts of writing. Attendees demonstrated this theme in different ways, so I will walk through some of my favorite interpretations⁴ across the four days of the conference.

Tim Brookes, director of the Endangered Alphabets Project, started the conference off strong with an exhilarating talk titled “The Betaverse”. In defense of the individuality of handwriting, Tim posed the idea that the concept of the alphabet — through its standardized representations of letters — disguises rigidity as uniformity, implying that our bodies can’t meet the standard forms that we have created. The betaverse, Tim proposes, is thus the antithesis to what the alphabet represents; it is writing as it is written by humans. To the beat of a live musical element, Tim whirled through a series of examples of these messy, chaotic, and organic handwritten forms that are representative of humanity. In addition to energizing the audience, Tim’s talk raised interesting questions: coming from a world of standards, the study of handwriting can perhaps offer a dose of healthy chaos in understanding the complete picture of writing systems.

Several of our SEI colleagues attended the conference, and it was interesting to see how they presented their work through the visual engagement lens. Oreen Yousuf, who works on African scripts with SEI, presented about Minim Dag Noore, a newly-invented script from Burkina Faso. Attendees were fascinated by the influence of Arabic and Latin on the script in a modern context — something the paleographers in the audience rarely get to examine. 2026 SEI research fellow Jordan Williams presented a portion of his research on Nsibidi, arguing that the user participates in the art of the signs, which creates semantic meaning. Daniel Yacob gave a fascinating talk about rubrication⁵ in Ethiopic manuscripts, which he described as a way for scribes to express themselves artistically and add another layer of meaning to the text.

Oreen presenting his talk “Look Both Ways – Minim Dag Noore”. The script’s bidirectionality and lack of mixed casing (i.e., the script is written entirely uppercase or entirely lowercase) make it an interesting case study for the visual engagement theme. Oreen explained that there doesn’t seem to be a mental barrier for people who read Minim Dag Noore right-to-left vs. left-to-right — the teachers grade homework written in the opposite direction of what they teach with no problems (Photo credit: Oreen Yousuf)

Jordan’s talk, “Public/Modern Nsibidi as Actionary Visual Experience”, allowed the audience to try interpreting the meaning of Nsibidi signs. Through audience participation, Jordan demonstrated that many people naturally connect to more visual cues. Nsibidi’s symbolic signs are thus quite versatile for representing actions and visual interpretations of the world

Daniel presenting “Color and Communication in the Ethiopic Tradition”. In Ethiopic manuscripts, the name of the Virgin Mary is often colored, even in (real) gold pigment. In hymns, the color red is used in musical notations to indicate performing with a somber tone. Due to the subtle semantic roles it plays, this raises interesting questions about the importance of color in Unicode. In theory, Daniel says that color fonts could be handled with the same technology used for emoji skin tones, or using variation selectors

Throughout the conference, several conversations concerning the active engagement of the user in the act of writing arose when discussing the encoding of historic scripts. Many talks at WAVE demonstrated that ancient scripts are closely linked to their physical mediums and the user’s physical interactions with the writing. When digitized, some of that information is lost in the move from the 3D world to a 2D computer screen — be that the act of engraving, the feeling of the physical medium, the layout of the text in space, etc. During the Liège workshop, Kaveh had echoed this concern as well, explaining that he had to figure out how to demonstrate deep vs. shallow impressions of Proto-Elamite text (originally inscribed on clay tablets) in his 2D font to preserve some of the physical context.

In a talk about Egyptian hieroglyphs, VIEWS research associate Jordan Miller proposed that ancient Egyptians may have engaged with writing more like folds rather than lines, since the glyphs would have been on rough papyrus or engraved in stone

Colton Siegmund, another VIEWS research associate, presented about examples of Sumerian inscriptions on statues of the ruler Gudea. These inscriptions, which contain gaps around the buttocks of the statue, are fundamentally tied to the physical context of the surface they are inscribed upon

Katarzyna Mikulska from the University of Warsaw demonstrated that the physical layout of a Mesoamerican script codex described in her presentation contains relevant semantic information for interpreting the meanings of the text. Additionally, the signs of physical wear and damage in the text suggest that the codex was used as a board/tabletop to perform divination rituals upon

Outside of the talks, the VIEWS team facilitated opportunities for us as attendees to engage with the Writing As Visual Engagement theme through activities. During Saturday’s lunch break, we tried out an exercise called “jamigraphy”⁶, which encouraged us to freewrite whatever letterforms came to us as we listened to music performed by Tim Brookes and Pule kaJanolintji. This was a nice way to connect with the artistic side of scripts, which I sometimes miss in the world of digital text encoding.

Conference attendees participating in the lunchtime “jamigraphy” writing activity

Conference attendees participating in the lunchtime “jamigraphy” writing activity

Conference attendees participating in the lunchtime “jamigraphy” writing activity

Pule and Tim providing the tunes for the “jamigraphy” session

Beautiful glyphs could be found all around the Faculty of Classics

My codepoint contribution to the “jamigraphy” poster

From the talks, social events, and even just the feeling of everyone sitting together in the audience, WAVE cultivated a strong sense of community that I really felt like I belonged to. Thank you to Pippa⁷, Tim, and the rest of the organizers for hosting such a welcoming, fulfilling event!

This was my first time meeting up with SEI research fellow Jordan Williams and SEI collaborators Oreen Yousuf and Adam Yeo in person (Photo credit: Oreen Yousuf)

WAVE reception in the Museum of Classical Archaeology in the Faculty of Classics

WAVE reception in the Museum of Classical Archaeology in the Faculty of Classics

The WAVE audience. The single-track format made this conference feel very close-knit

A lot of the value from these in-person conferences comes from the social opportunities outside of the talks

Group photo of attendees at the Writing As Visual Engagement conference in Cambridge, UK (Photo credit: Pippa Steele)

All sorts of fascinating goodies brought by WAVE attendees

Adam Yeo with his book on Bété

Saying goodbye to the beautiful Faculty of Classics at the University of Cambridge

Overall, I had a wonderful time observing writing systems from a more academic perspective and meeting up with so many friends of SEI in Liège and Cambridge. These are rare and cherished opportunities in this work, so it feels extra rewarding to make the most of them like this.

1. A type of hieroglyph composition in which individual signs are arranged in a block formation. ↩︎
2. A type of hieroglyph composition in which individual signs are enclosed in an oval-shaped outline. In Egyptian hieroglyphs, this indicates that the enclosed text is a royal name. ↩︎
3. Anushah Hossain delves into the chicken-and-egg problem and the implications for encodings in “Unsettled Scripts” available here. ↩︎
4. Many of the talks from WAVE 2 are available online here. ↩︎
5. The process of scribes making text red to make it stand out in some way. ↩︎
6. Read more about “jamigraphy” on the VIEWS Blog. ↩︎
7. Read Pippa’s wonderful reflection on the VIEWS Blog about WAVE 2. ↩︎