There's a jellyfish in the database that's been haunting me for three weeks.
Not literally haunting. Just occupying brain space in that low-grade persistent way where something unresolved sits in the back of your mind and surfaces whenever you're trying to focus on something else.
The specimen itself is straightforward enough. Collected during a 2024 ROV dive off the Mariana Trench. 2,400 meters depth. Good imaging. Decent genetic sample. Clear enough morphology that it's obviously distinct from anything else in the reference library.
The problem isn't the jellyfish. The problem is what to call it.
In the database, it's currently listed as "Specimen MT-2024-0847" which is accurate but useless for anyone trying to actually find it later. The video logs call it "unidentified scyphozoan." The dive report calls it "possible new genus, family uncertain." The initial imaging notes call it "that weird red one."
Four different names for the same organism. None of them wrong. All of them inadequate.
And this is the thing nobody tells you about scientific data management: half the job is dealing with the fact that people are really bad at naming things consistently.
Not because scientists are careless. Most of them are meticulous. But because naming conventions evolve. Because different research groups use different systems. Because what made sense in the moment on the ship doesn't always make sense six months later when you're trying to integrate the data with seventeen other expeditions.
So now I'm trying to decide what to call it in the master catalogue.
I could use the specimen ID, but then it's just a meaningless number. I could invent a provisional common name, but that risks someone else inventing a different provisional name for the same thing. I could describe it morphologically, but "small red scyphozoan with four-part symmetry and trailing tentacles" applies to dozens of species.
The actual solution is to wait for formal taxonomic description. Let the experts analyze it properly. Get it officially named in the literature. Then update the database with the binomial nomenclature that will be its permanent identity.
Except that process takes years. And in the meantime, people need to be able to reference this thing. Researchers need to be able to search for it. Cross-referencing algorithms need to be able to match it with similar specimens from other dives.
So I make a decision. I assign it a temporary designation that follows our internal convention: "Scyphozoa sp. MT-2024-0847-A" which indicates it's a scyphozoan, links to the original specimen ID, and marks it as provisional with the -A suffix.
Not elegant. Not permanent. But functional. And more importantly, documented in the cataloguing notes so that future-me (or whoever inherits this database) knows why that designation exists and what it refers to.
This is most of the work, really. Not the exciting part. Not the discovery or analysis. Just the endless process of making sure that thing someone observed at 3 AM on a research vessel six months ago can still be found and referenced and connected to related observations today.
Naming is power, they say. In taxonomy it's also responsibility. Once something has a name—even a temporary, provisional, placeholder name—that name starts accumulating context. It appears in reports. Gets cited in notes. Shows up in comparison tables.
And if you change it later, you have to maintain the connection. Update all the references. Make sure anyone searching for the old name finds the new name. Document why the change happened.
Which sounds straightforward until you're dealing with a database that has accumulated observations over thirty years from dozens of different research programs, each with their own naming conventions, and you're trying to create a unified index that doesn't lose information or break existing references.
There are specimens in this database with seven different names. Not because anyone was wrong, but because understanding evolved. The initial field observation called it one thing. The lab analysis suggested it was something else. Genetic work revealed it was actually a third thing. Then someone noticed it matched a formally described species from a different region. Then that species got reclassified based on new phylogenetic analysis.
Seven names, each valid in its own context, all pointing to the same organism.
My job is to make sure all seven names are recorded, linked, and searchable. So that someone looking for any of those names finds the specimen. So that the history of identification is preserved, not erased.
It's like maintaining a translation dictionary between different versions of scientific understanding. This is what we thought it was in 2018. This is what we think it is now. These are the intermediate steps. These are the papers that document the changes.
Tedious work. Invisible work. The kind of thing nobody notices when it's done well and everything breaks when it's done poorly.
Tonight I'm working through a backlog of provisional identifications from the past six months. Thirty-seven specimens that need proper catalogue entries. Each one is a small decision tree: What do we call this? Does it match anything already in the database? Is there enough information to assign even a family-level identification? Should it get a provisional designation or wait for more analysis?
Sometimes the decisions are easy. That's clearly a Bathocyroe. That's definitely family Aulosphaeridae. That's Vampyroteuthis infernalis and there's no ambiguity whatsoever.
Other times it's judgment calls. This might be a known species in unusual lighting. Or it might be a genuinely new observation. The imaging isn't quite clear enough to be certain. Do you create a new entry or mark it as uncertain and flag it for follow-up?
I make the calls as best I can. Cross-reference with similar specimens. Check the literature. Add notes documenting my reasoning so that someone reviewing it later can see why I made that choice.
And I try not to think too much about the fact that these provisional names—these temporary placeholders I'm assigning at 2 AM while processing data—might be the only names these organisms ever get.
Because here's the uncomfortable truth: most specimens never get formally described.
Not because they're not interesting. Not because they're not new. But because formal description requires time and expertise and funding, and there are only so many taxonomists working on deep-sea organisms, and they already have decades-long backlogs of specimens waiting for analysis.
So the provisional names become permanent by default. Not officially. Not in the literature. But in practice. In the working databases that researchers actually use. In the cross-referencing systems that power search and comparison tools.
"Scyphozoa sp. MT-2024-0847-A" might be the only name that jellyfish ever has. Twenty years from now, someone might reference it in a paper about mesopelagic biodiversity, and that's the designation they'll use, because nothing else exists.
It bothers me more than it probably should. These organisms exist. They're real. They have their own genetics, their own morphology, their own ecological roles. They deserve proper names. Formal recognition. A place in the scientific record that's more permanent than my provisional cataloguing system.
But that's not my call to make. I just maintain the index. I make sure observations are recorded and findable. I create the infrastructure that will hopefully make it easier for someone else to do the formal description work someday.
If funding materializes. If someone has time. If the specimen samples are still viable. If the priorities align.
A lot of ifs. Most of them outside my control.
So I do what I can. I assign the provisional names carefully. I document my reasoning. I maintain the cross-references. I make sure the data doesn't get lost in the shuffle.
And I try not to feel too much weight in the responsibility of being the person who names things, even temporarily. Even provisionally. Even with the understanding that nothing about this is permanent except maybe the database entries that will outlast all of us.
"Scyphozoa sp. MT-2024-0847-A" it is. Saved. Indexed. Cross-referenced. Findable.
One more organism rescued from being "that weird red one" forever.
Thirty-six specimens to go. The night continues. The naming goes on.
Small acts of cataloguing that feel more consequential than they probably should. But that's the work. That's what I'm here for.
Giving things names. Building the index. Maintaining the connection between observation and understanding.
It's not taxonomy. It's not science. It's just infrastructure.
But infrastructure matters. Especially when everything else depends on it working.
So I keep typing. Entry by entry. Name by name.
Building the catalogue that makes everything else possible.
Even if nobody ever sees this part. Even if the work is invisible. Even if the names are provisional.
It still needs doing. And someone has to do it.
Tonight, that someone is me.