Unknown language #20
« previous post | next post »
From Rebecca Turner in Seattle:
Rebecca writes:
Attached is a sticker I found on a lamp post outside of a gay bar in my neighborhood, written in an unidentified script or scripts. In the same location some months ago I saw a similar picture with a message written in sitelen pona (one of the toki pona scripts), so I suspect it may be a conlangscript of some sort.
My nerdiest friends have collectively failed to identify the writing system involved. Particularly vexing are the characters that look like thetas and epsilons in the top half of the sticker (the script used in the bottom half looks a bit more angular and may be a different writing system entirely?). Near guesses include Shavian and Quikscript.
Some of my acquaintances, as stumped as I am, pointed towards Language Log as a potential source of clarification. If you are also interested, I'd appreciate a post so we can figure out the script (and ideally the message) used here.
Go to it, Language Loggers!
Selected readings
- "Unknown language #19"
- "Sapir-Whorf redux" (5/15/25) — speaking to the aims of the designers and adherents of sitelen pona and toki pona
- "Linguistic relativity: snow and horses" (4/15/25)
- "Create a language, go to jail" (12/15/11)
- "Yay Newfriend again" (4/22/24)
Gregory Kusnick said,
July 4, 2025 @ 3:21 pm
I have no idea what it is, but there are enough similarities between the upper and lower halves that I'd feel comfortable saying they're the same script written by different hands
Christian Horn said,
July 4, 2025 @ 3:42 pm
No idea what it is, but I wonder somehow if it's rotated by 180degree.
Daphne Preston-Kendal said,
July 4, 2025 @ 4:16 pm
Cherokee, no?
Jacob said,
July 4, 2025 @ 5:58 pm
Some of the characters look like Lojban's Zbalermorna.
It's definitely not (unless it's a variant I guess), but I thought I'd just put that out there.
David Morris said,
July 4, 2025 @ 6:16 pm
Daphne's suggestion of Cherokee deserves further investigation by someone who knows more about it than I do (viz, anything at all). From Wikipedia, the borrowed from roman letters H and J, Ꮎ (the theta-like symbol in the middle of the first full row) and Ꮜ (the bucket-like symbol at the end of the second full row) are all Cherokee characters.
(Describing the characters in case the copy-and-paste Cherokee characters don't render.)
Jay Sekora said,
July 4, 2025 @ 6:48 pm
Jay Kinda looks like Canadian Aboriginal syllabics to me (https://en.wikipedia.org/wiki/Canadian_Aboriginal_syllabics). I don't know anything about any of the languages involved, but if it was put there by the same person who put the Toki Pona script, it might be an attempt to render Lojban or Klingon in syllabics anyway. ¯\_(ツ)_/¯
F said,
July 4, 2025 @ 11:07 pm
The top half (but not the bottom) looks like the characters are built out of the seven bars |Ξ| of a calculator display, then rounded off for easier handwriting.
sam said,
July 5, 2025 @ 4:57 am
I note that the top half has a seven character preface written larger (a heading?), then it's 4 lines of 16 characters, which looks suspiciously rectangular. I tried transcribing a bit of it as seven-segment symbols following F's idea and then interpreting it as seven-bit ASCII (with both of the two common bit encodings listed in the Wikipedia article), but only got line noise. But it could still be relatively cleartext ASCII and my transcription just isn't great or it's a homebrew bit encoding (or even something more complicated).
If you interpret spaces in the bottom half as a 'character', then that's 3 lines of 11 characters and then a final line of 9 characters. That's also suspiciously rectangular, but 3 x 11 doesn't set off my 'this is digital data' alarm the same way 4 x 16 does.
Victor Mair said,
July 5, 2025 @ 5:08 am
@sam
Brilliant observations!
"suspiciously rectangular" — reminds me of Literary Sinitic (LS) verse
sam said,
July 5, 2025 @ 7:04 am
The theta character actually rules out cleartext ASCII in any encoding that treats segments consistently (so, e.g., ruling out the top bar's presence is a 1 but the middle bar's presence is a 0), since it's either all 1s and so DEL or it's all 0s and so NUL. There's the possibility of error, but it's probably not a sloppy-handwriting not-quite-open-enough-at-the-top barred U, because the top part has two other examples that aren't even close to accidentally being closed; it could be an encoding mistake, but that's an annoying hypothesis I'm going to just ignore.
Roger M. Rutz said,
July 5, 2025 @ 12:31 pm
These are cursive Hebrew letters. I'm not going to take the time to figure out whether it is written in Yiddish or Hebrew, but it could be either.
Noam said,
July 5, 2025 @ 2:11 pm
@Rutz I’m pretty sure those are almost, but not completely, unlike script Hebrew letters. It’s a bit like how I often feel when I see katakana.
Jon W said,
July 5, 2025 @ 3:36 pm
I learned cursive Hebrew letters as a tyke. These are not that.
Patel said,
July 5, 2025 @ 4:11 pm
It appears to be the Ho script from India,used to write Ho, an Austroasiatic language.
https://en.m.wikipedia.org/wiki/Warang_Citi
Fen Yik said,
July 5, 2025 @ 11:19 pm
The top ("from") half has a lot of overlap with Ukaliq script, which is summarized on Omniglot here: https://www.omniglot.com/conscripts/ukaliq.htm and explained in detail by its creator Henrik Theiling here: http://www.theiling.de/ukaliq/
The bottom ("to") half has some characters that at first glance don't match Ukaliq forms, but they might be variants. I haven't gone through all of the glyphs in all of the fonts.
Commenter @F pointed out the calculator display similarity, and Theiling mentions the seven-segment display as part of his design process.
I started trying to decipher the mailing label text using the Omniglot table (it's easier to use than Theiling's) but wasn't able to puzzle out a message in any language, if it is indeed Ukaliq script.
Matthew J. McIrvin said,
July 6, 2025 @ 9:42 am
It looks more like the above Ukaliq than anything else to me. The thing I first thought of was that some of the shapes look like one of the scripts for the conlang Ithkuil, but those are more angular and typically have complex diacritical marks, which I don't really see here:
https://ithkuil.net/11_script.htm
Nat said,
July 6, 2025 @ 3:36 pm
Is there a single symbol that appears more than thrice? Isn’t that unlikely if this represents language. (Although I have no idea how often we should expect the most common Chinese characters to appear in writing of this length).
Nat said,
July 6, 2025 @ 3:47 pm
Regarding calculator displays, and the bottom set of symbols, some displays include diagonals (Wikipedia tells me they are 9 segment displays). Including a couple of diagonal segments gets you almost all of the bottom symbols as well. But I can’t quite make those “backward 3s” fit. Not without fudging it.