Text Is Data

Approaching Heritage Translation as Digital Curation

What Kind of Translator I Am

When people hear the word “translator,” many picture someone who translates novels or poetry: a person who works to preserve the texture of the original, weighs the rhythm of each sentence, and might spend days on a single word. I am not that kind of translator. I don’t translate literature, and I don’t approach translation as an artistic pursuit.

My work is to convey Korea’s cultural heritage to international readers accurately and in a way that is easy to understand. Most of what I do is heritage site signage, museum exhibition text, educational materials, and abstracts for academic journals. And I do this work not only as a translator but as a digital curator. I studied cultural informatics in my master’s and doctoral programs, and I worked as a heritage translator throughout my studies. Over the past ten years I have handled more than 6,000 heritage sign texts, and together with colleagues at the Academy of Korean Studies, I contributed to the Korea Heritage Service’s guidelines for heritage signage.

Where these two backgrounds meet, I came to see text a little differently. To me, before a single sign is a piece of writing, it is one piece of data that needs to connect to thousands of other texts.

A Name Is Data

An explanatory text printed on a gallery wall does not end there. Visitors search the names they see on signs, and they ask AI about them. Institutions post the same text on their websites, enter it into databases, and someday hope to link it with the materials of other institutions.

What happens, then, when the same heritage site is written with a different English name at each institution? Visitors cannot tell that it is the same place, and a search cannot connect the two. If the same thing has been named five different ways, someone is left with the job of linking those five names together. And because AI learns from published text, the more scattered our terminology is, the less accurate AI’s answers about Korean heritage become.

This is why I ask “Is everyone using the same term?” before I ask “Which translation is the most correct?” I am not against the search for better translations. But as long as a term is not wrong, having one unified term is worth far more than having several scattered ones. This is the reason I follow the government’s standards first: the Romanization of Korean, the guidelines for translating public terms, and the Korea Heritage Service’s rules for the English names of national heritage. They matter not because they are perfect but because they are a single standard that everyone can follow together.

Following guidelines can look trivial. Is there a space before the parenthesis? What format does the name of an inscribed object take? But from the perspective of data, these details are everything. To a computer, a name that differs by one character is a different name entirely.

Why Guidelines Are Hard to Apply

The standards already exist, and they are well made. Yet in the English texts actually in use, I often come across names and terms that do not match the existing guidelines. As someone who works alongside these institutions, I do not believe this comes from anyone’s carelessness. As I see it, the reasons are mostly structural.

First, most staff at public institutions work on a rotation system. Just as someone becomes familiar with the English standards, they are moved to another department, and their successor starts again from the beginning. No one who remembers the standards stays in place for long.

Second, translation is usually outsourced, and it is only natural that the staff member in charge is not an English specialist. This means the institution has no internal way to check whether the text it receives meets the standards.

Third, Korean texts go through a careful process of refinement, including review by the National Institute of Korean Language. For English texts, in most cases, no equivalent step exists. Translation easily becomes the last stage of a project, handled with whatever time and budget remain once the main work is done.

Fourth, the larger a project is, the more it is divided among multiple companies and translators over multiple years. Without a shared glossary, the same term gets translated in different ways, and words that could have been expressed in English are left in Romanization.

Under these conditions, each institution is doing its best with the resources it has. But the result is that there is no one, at any point in the process, whose job is to check the text against the standards.

What This Looks Like in Practice

Here is one example. According to the Korea Heritage Service’s naming rules, the name of an object bearing an inscription takes the format “with Inscription of.” As of this writing in September 2026, however, the English name of a bronze gong designated in 2024 is registered as follows:

Bronze Gong inscribed with “Cheonsuwon” (Cheonsuwon Temple)

Following the rules, it would be:

Bronze Gong with Inscription of “Cheonsuwon”

This may not look like much of a difference. But when objects of the same type are registered in different formats, they can no longer be found with a single search or handled by a single rule. Similarly, in the English name of an item of intangible heritage designated in 2026, “Multtae(Knowledge of tide),” the missing space before the parenthesis and the capitalization inside it both depart from the standard.

I do not give these examples to blame anyone. If this can happen at the very institution that wrote the standards, then it is not an individual’s mistake but a gap in the process. Had there been a step where each new name was compared against the rules before being registered, these would have been caught. That step simply did not exist.

What I Do

I want to be that step: the person who still remembers the standards after the staff have rotated, and who can check the text you receive against them.

In concrete terms, this is what I do.

Translation and review based on the standards. I follow the government standards and The Chicago Manual of Style, and I check that official names also comply with the naming rules. Where something does not match, I let you know, along with the basis for the correction.
Institutional style guides and glossaries. I put an institution’s standards into a written document so that consistency holds even when many translators work over many years. Standards that exist in writing remain when the staff change.
Consistency checks for large bodies of text. I treat hundreds or thousands of texts as data to find where terms and formats diverge. This is where the methods of digital humanities are put to the most direct use.
Design proof review. Before printing, I check for the copy-and-paste errors, missing italics, and typesetting problems that arise when a translation is moved into a design file.

The detailed standards I follow in my own work are published in my Working Standards.

A Few Suggestions

Finally, I would like to offer a few suggestions to institutions that produce English texts.

I hope English texts can be given the same weight as Korean ones. To an international visitor, the English text is the face of the institution and, by extension, the face of Korea. The English texts of government agencies and national and public museums in particular become a reference that other institutions and translators follow, so their influence does not stop at the institution’s walls.

When beginning a large translation project, I recommend deciding on the standards first. Choosing which style guide and which glossary to follow, and then maintaining consistency by those standards, creates value that lasts longer than the sheer amount of text produced.

When selecting a translator, one criterion you can use is whether they know and follow the government’s basic standards. This does not tell you everything about the quality of their work, but it does show how well they keep up with current standards and how much attention they pay to detail.

I also believe it would be good for the translator of a public institution’s text to be named in the final product. These texts are made with public funds, and many people trust and refer to them. Putting one’s name to the work and taking responsibility for its quality is the translator’s part, and I intend to begin with myself.

Closing

In translation, I think about accurate and consistent data before I think about beautiful sentences. I believe this is how Korea’s cultural heritage will come to be properly searched, connected, and understood in the digital environment. If this perspective resonates with you, or if you have been working through similar questions, please feel free to reach out through my contact page.

September 2026