Earlier during the pandemic fun I suggested that simple and easy labels be applied to the variants like: Variant A, Variant B, etc. Although it’s not exactly analogous this is what we do for hepatitis (Hep A, Hep B, etc) and some other maladies. Seems like a good solution to me but I suppose it’s too late.
What is too late? What is stopping you from using “real” names like B.1.427 and 20C/S:452R which carry a lot of actual information about the evolutionary history instead of names like “California variant”, which, ok, tell you where it was first detected, but nothing more?
ETA
that’s what the Nextstrain labels do. Except it’s not A and B, you have 19A, 19B, 20A, 20B, 20C, etc., up to 21A now (the date is included)
I’d never heard of Nexstrain. I wonder why those simple terms aren’t being more widely used?
I don’t know why they do not use it either, but here you go, the information is hardly kept under wraps, quite the opposite:
https://nextstrain.org/ncov/global
variant info:
https://www.cdc.gov/coronavirus/2019-ncov/cases-updates/variant-surveillance/variant-info.html
These “India variants” seem to be 20A, for example.
Now that I’ve looked at Nextstrain more closely, I’m not sure the variant names there are any more helpful.
From what I could tell only the early “parent” variants had simple names like 20A or 19B whereas a more current variant has a name like 20I/501Y.V1
Are there any other current 20I subclades besides 501.Y/V1? I assumed that was originally a 20B (or some other) variant that got assigned the name 20I. As per the instructions:
I.e., the “major” variants really are supposed to have a short name consisting of the year +letter.
Please also note the names on the Pango website, starting from lineages A and B, branching down to A.1, A.2, B.1, B.2, etc., and more sublevels like B.1.1, B.1.2, etc., except that a maximum of three sublevels are allowed, so (hypothetically) B.1.1.1.1 would become C.1, etc.
ISTM both Pango and Nextstrain are absolutely hewing to your “letters” principle, more or less, grosso modo under the constraint that there are lots and lots of branches so you need a little more than merely going from A to Z.
I have not thought about it too much myself, but if you have an obviously superior idea then I am sure they would be happy to adopt it 
The source code for e.g. Pangolin GitHub - cov-lineages/pangolin: Software package for assigning SARS-CoV-2 genome sequences to global lineages. is available in case anyone wants to see precisely how the names are currently generated.
Apparently India isn’t fond of the term.
That article says that the WHO’s “new” guidelines include a scheme to name virus variants analogous to the way tropical cyclones are named (because the public is obviously too stupid to comprehend Pango and Nextstrain names like P.2 and 20I??) but does not give any concrete examples. Not that it’s too complicated to turn P.2 into “Priscilla” or go with nonsense syllables like “Paidorin”.
Do you really believe the public at large can remember a half-dozen or so - or more - of these styles of names and keep them straight and talk meaningfully about them? I don’t, and seemingly neither does the mainstream press. That’s why they use the geographic names.
It’s sweet of you to say that. I only wish you’d said you thought my opinions were silly. I could put that in a sig.
It is not quite accurate to say they simply “use the geographic names”. For example, one article starts off with “A coronavirus variant first identified in India is now officially ‘a variant of concern’” but goes on to repeatedly, and exclusively, refer to it as B.1.617.
Another article, still also ante-dating the “new” WHO convention, whatever that might turn out to be, suggests there are so many variants that people can’t keep all the artificial names straight. Examples given were “Pelican”, “Quail”, “Mockingbird”, “Kingfisher”, “Pigeon”, “Doug”, “Douglas” (not to be confused with Doug!!!), “Nelly”, “Eeek (aka Eric)”. Try to remember which is which six months later.
You know, you’re right, and I overstated that. Some, though not all, media outlets try to split the difference by referring to the variants by their genetic code, with a single reference to the more common geographic name on first use. Personally, I still find this hard to parse, and make a couple mental notes of which code maps to which variant (for which I have some pre-existing knowledge [eg, UK Variant = more virulent, but seems to be fully blocked by vaccine; SA Variant = maybe sneaks through the vaccine a bit more, etc]). I can keep those in my head for the duration of the article.
I do wish we had a convention more like this one you mention:
Sure, that’s not going to scale forever. If there are 200 variants, no one’s going to remember those either. But for ten or so, that would work where the code naming scheme won’t.
It also doesn’t really convey any useful info about the virus. In the case of “China virus,” there are many, many viruses that were first identified in China. Even “Wuhan virus” wouldn’t have been specific enough: there are something like a dozen viruses that were first identified there. And knowing the place of origin isn’t particularly useful once the virus has spread beyond its geographic origin.
And so it goes with the variants. As soon as we ID a second or third variant in India or some other country, the nomenclature will get confusing.
This is pretty much it. Early last year the media was calling this new virus “China/Wuhan virus,” and then in March the WHO followed their 2015 naming guidelines and gave it the official name “SARS-COV-2”, and they named the illness caused by this virus “COVID-19”. The media (and the public) don’t like long complicated names, so “COVID-19” got shortened to just “COVID,” and “SARS-COV-2” just got shortened to “coronavirus” (even though there are many different coronaviruses).
Frustratingly, we (and the media) haven’t found a similarly catchy shorthand for the variants that doesn’t invoke their geographic origins. Until someone does, the media will keep reinforcing the use of placenames to identify variants - whether it’s “* variant”, or “variant first identified in *”.
You’re right; it doesn’t. It’s just a tag that gets associated. But I disagree that there’s a real ambiguity problem: we all know what West Nile virus is (well, many of us do.) It doesn’t matter how many viruses emerged along the west Nile; only one is West Nile virus. Same with Ebola and Spanish Flu.
Long before I ever heard “SARS-CoV-2”, I recall we were calling it “the novel coronavirus.” I think that’s the usage that just got shortened to “the coronavirus” or “'rona.”
“B1-617” is pretty easy for people to get used to, just like COVID-19. There’s no reason to give racists an excuse to target another Asian group.
If it were the only one, sure. But, quick, which one is the UK variant? How about the South African variant? Brazil variant? If you know those all, congrats. I don’t. I’d have to look them up.
And India is a big place with a large population. It could develop more than one variant pretty easily, I would think. Then what? Referring to the country doesn’t guarantee any clarity either.
Why would you need or want to look stuff like that up? The whole point is that we (or, at least, the WHO) agree that correct terminology does not include potentially misleading “[geographical] variant”. For example, which is the California Variant? B.1.427 or B.1.429?
and also
The geo name just indicates where a particular sample was isolated, and, again, there are thousands of these.