NULL Gender

It was October 2023. I was sitting in front of a desk at an immigration office. The lights were fluorescent, and the office was small: just enough space for the desk and some chairs. Two other asylum seekers were sitting beside me. I was located against the wall, then one in the middle, and the last one near the door. We were silent, uneasily waiting for the officer to arrive. After a few minutes, the officer came in. She had a friendly and respectful disposition, and she was carrying three thick envelopes, full of paperwork.

She began reviewing forms and information with the person closest to the door. She asked them to check their name, to double-check their address, and to see if all other data points were correct. They were. The officer asked them to sign some papers and, in the end, gave them the whole envelope. It included their Refugee Protection Claimant Document, their work permit (valid for two years), as well as copies of their refuge claim, general information forms and some manuals and guidelines. The refuge claimant said thank you, stood up and left the room.

Then it was the turn of the person sitting next to me. It all went in a similar fashion. They checked their information, double-checked their address, signed some papers and got back a similarly thick envelope full of copies, IDs, permits, forms, and manuals. They also said thank you and left the room.

It was then my turn. I expected things to go exactly the same way. I was asked to check my name, double-check my address, and verify all the rest of the data elements that conformed my identity on paper, but then there was one big difference, I was offered an extra form, one that the other two asylum seekers weren’t offered: a form to legally change my gender marker.

As a transgender person, there are some milestones I look forward to. A big one is when I will finally be able to change my legal name, but that one has to wait, lest it interfere with my still ongoing immigration process. Another one was the day when I would legally changing my gender marker.

I had thought of that day for a long time. How I would be preparing for it, how I would be so filled with anticipation, and how it would be one of those big days you remember for the rest of your life. I was planning on making it special, memorable not only because I changed my gender, but also for all the ritual I would build around it.

Turns out life, being as unexpected as it always is, had something else in store for me. It was in this small, fluorescent-lit office room where all of a sudden I had the paper in front of me. Life is funny; we try to make it grand, and it hits us back with mundanity. However, even if it was not with the fanfare I expected, I was prepared for this event. I knew this decision was important in many ways other than the evident.

I already knew I would have three options: keep M, and leave the decision for later; choose F, and settle on the other side of the binary; or pick the new X, and express my nonbinary nature for the world to see. Like I said, I had prepared for this moment: no matter how much I dislike binaries, I chose F.

I grew up in Costa Rica in the 1980s, where being transgender wasn’t even a thing. There you were a transvestite, and generally seen as a default sex worker. Culture hasn’t changed much over these last few decades. During the late 1990s and early 2000s, I attended university as an apparently cis male and had the privilege of studying computer science, a particularly male-dominated career, without much of a fuss.

There were courses about programming, operative systems, computer architecture, quality standards, collaborative systems, and lots more. However, one of the most expansive areas of knowledge was data management. How databases work in all dimensions, from physical storage, infrastructure, data-center organization to more abstract subjects such as data modelling, optimization, and structure. One particularly important subject was database design: how to store information in the most efficient and cost-effective way possible.

One of the first things you learn about data is that it is costly. It costs money to pay for storage, electricity, and access, and it also costs processing time reading and writing it. When you design databases, you are taught to put efficiency as your top priority. Even if the difference is fractions of a second, when you read millions of data elements, it adds up.

Data is stored in Bits: what is commonly known as 1s and 0s. It is the smallest and shortest data format possible in the binary-based computers we most commonly use, and can represent a value that can be either a zero or a one. Of course, storing only two possible values doesn’t allow for much data expression, so there are other types of elements. Bits can be grouped in Bytes (eight Bits together), which allows us to represent numbers from 0 to 255; and Bytes can be grouped together, representing numbers from 0 to 65,535 using two Bytes, 0 to 16,777,215 using three, and so on. There are also strategies for managing fractions and more complex numbers.

Storing text is somewhat more complicated. Since computers can only handle numbers, the solution of choice was to assign a numerical representation to each letter. So, in the early 1960s the Washington D.C.-based American National Standards Institute (ANSI) approved the American Standard Code for Information Interchange (ASCII). ASCII is one of the most generalized ways, still in use, to represent text in computers and networks. Using it, every letter (lowercase and uppercase), number and symbol in the English language — or at least those the people designing the standard thought of — has a corresponding numerical code.

Since computers store everything in binary, those numbers end up being quite long. So, the letter “A” is represented by the number 65 and stored as 1001101, the letter “a”, by 97 and stored as 1100001, an “@“ symbol is 64 or 110100, a question mark is 63 or 110011; so on and so forth until all letters and symbols are covered.

There are glaring problems with this list. For example, it doesn’t include letters that don’t exist in the English language, such as “ñ”, “Ç”, “Æ”, and so many others. To solve that, other encoding formats have been proposed, such as Extended ASCII, Unicode, and UTF that, in very few words, make the numbers representing each character longer in order to have more options. The most recent ones cover characters in most languages, such as Chinese ideograms, Cyrillic, Kanji or mathematical symbols like ∑, which is represented by the number 8721, or 10001000010001. Things can get really long really fast. And they use more and more storage space and processing time.

These are all very clever approaches for representing information, but there is one additional situation: what happens if the value is unknown? For example, if you don’t know someone’s last name, their address, their date of birth, or their height or weight. For those cases, you use NULL. A special value that explicitly states that data element has no value. Because it’s not the same saying someone weighs 0 pounds than saying their weight is unknown.

So, all of this is to say that databases are complicated, and optimization is paramount. If you need to store a data element that you know is only ever going to have two possible values, you use a data type that has only two possible values: a Bit. No need to waste storage and processing power allowing more options, if you know that data element will never be other than X or Y. Or M or F. You assign 0 to F, 1 to M and you are done. You even make it read-only (which makes it more efficient to access) because you also know it is never going to change.

If you were a cis white male designing a database in the 1950s you most likely would have seen gender as binary and unchangeable. It just made sense. It’s a simple, efficient and, dare I say, elegant way to represent everybody’s gender. No need to involve longer numbers or — shudder — letters.

Welcome to hard-coded, binary and unchangeable gender markers.

So, no matter how much I dislike binaries, I chose F.

I knew gender is generally stored as a Bit; a binary value. I knew I wasn’t choosing F, but either 0 or 1, depending on how that particular database was mapped.

But then, why not X? As I mentioned, I knew there were three options: M, F and X. X is the recent revolutionary adaptation made in computer systems to represent gender outside of the binary, right? Something that is not M or F. In fact, how is it actually possible to have a third option if a Bit can only be a one or a zero?

This is the happy scenario: computer engineers have already painstakingly modified those old systems to accommodate for different gender identities; they implemented X, breaking out of binaries and Bits. Reality, as it turns out, is never that simple. X doesn’t represent nonbinary; X represents NULL. If you only have a Bit for a data element, you only have 0 or 1, but you can also remove its value and leave it unassigned. This is not a value, it is the erasure of all values. X is nonexistence of information.

I chose F because I don’t want my gender to be erased, unassigned, undefined or nullified. I want it to be represented. I am not satisfied with F, but the solution isn’t accepting implicit erasure. The solution is designing systems that regard people as more than simplistic data points. Where their demographic information goes beyond predefined lists of numbers and turns into everchanging stories. If that makes systems inefficient, then we have to find ways to improve them, not to simplify people in order to satisfy them.

It may be a gargantuan effort, but it can be done. There have been previous cases where data models have proven to be faulty and they have been modified accordingly. Take the Y2K problem for example (nationalgeographic.org), where the year number on dates was stored only as its last two digits; when the year 2000 loomed near, companies rushed to fix it because it could cost them millions.

However, when a system obscures the true complexity of gender identity, when they promote outdated and arbitrary gender binaries, it is the people who are misgendered who bear the cost. Organizations aren’t motivated to invest in improving their systems for them, because they don’t feel any tangible impact. It is up to us to find a way to pressure them, to find alternatives and to promote change.