- Write down a bunch of words vaguely associated with the project.
- Scrabble their first letters to get words or word-ish sequences that are at least pronounceable.
- Try to ensure that the sequence of words is at least not objectionable.
And that's how we end up with a name like PACEM (Proto / Amorphous Collective Energy Management) or TASBE (a Toolchain to Accelerate Synthetic Biology Engineering). But honestly, those are just not terribly good names, because I don't have a natural knack for branding and I haven't ever really invested in it. Now "Scrabble", that's a name that I can really get behind: it really conveys the scrambling and shuffling that goes on when you play the game, but doesn't go acronymic to take the lazy way out.
Naming is on my mind today, because we've been doing a huge overhaul of our Proto programming framework this summer, and since the programming language that comes out of this will look very little like Proto currently does (for those who care, it will switch from prefix to infix notation and all those LISP-ish parentheses will go away) we might take the opportunity to coin a new name. So we've been kicking around ideas and trying to get away from the acronym thing.
It's really tempting to get cute when you're naming things, and in my experience this always will tend to end badly. The problem is that once you make a name, you get stuck with it. A joke is nice, but do you really want it pinned on your lapel for a decade of your career? Or worse, a joke can end up with you not able to be found at all, because you get lost in the noise or incomprehension. Though, perhaps not always... I had an undergraduate working for me some years ago, who came up with an extremely fast and elegant algorithm for accelerating complex chemical simulations. When we were nearly done writing it up, I told him that this algorithm needed to have a name, and he asked me, "What are the rules for naming algorithms? Can I name it anything I want?" Incautiously, I told him yes, having apparently forgotten the madness of power that would have overtaken me at these words when I too was an undergrad.
And so it is that I came to be coauthor on a paper about the LOLCAT method for chemical simulations. I could have quashed it and demanded we have a better name, of course, but he was so happy with the idea and it was really mostly his invention, and so I gave in and we've got this crazy name. It's not even an acronym with some sort of reasonable excuse, it's just LOLCAT because my undergrad was mad with power and hilarity, and I was quite content to let him have his fun. If you google LOLCAT, well... either you already know, or you should do it now. If you google "LOLCAT algorithm," though, we come up on top, and (I just discovered) it's even apparently gotten me into Wikipedia. So I guess that means that names don't matter too much after all, if you've got something useful that you've built, in the end.
I'll definitely say, though, that there's one thing I'm very proud of in my history of naming: Harriet (seen here in her mother's lap, watching airplanes landing over her into Logan)
That name was subject to extremely careful thought, since a person would be branded with it for her whole life (or at least until she gets old enough to change it legally if she chooses). It came to me one night about halfway through my wife's pregnancy, as I was biking across town and tumbling around my thoughts on naming constraints: instantly recognizable but not too common, no strong cultural baggage from famous people or stereotypes, needs to have good "weight" on the tongue, something that can enable a strong woman to choose whoever she wants to be when she grows up and not get in the way of her being taken seriously. So far, it seems, that Harriet has fit the bill, and she wears it proudly, along with her Indian middle name, Purna. We'll see which one she settles on as she grows...