Can you read Uzbek, Croation or Serbian script?

April 13, 2023 2:17 AM

I have some postcards from the WW1 era with short handwritten messages on them in a language I can't read. Google Translate suggests it may be Uzbek, with a caveat that Croation or Serbian is also possible. Because they're handwritten in cursive and I have no context in any of these languages to make intelligent guesses, I can't enter more than a word or two into a computer translator, so I'm hoping to find some human help instead. If that's you, please add a note below or Metamail me. Thank you.

The cards are from the same batch we've already discussed here. Like the coded card, I think these foreign ones were written by a teenage boarding school pupil to one of his older relatives at home. There's four or five cards in all, none containing a message of more than about 150 words. I'm hoping someone may be intrigued enough to do this just for the fun of it, but willing to agree a small payment for the job if necessary. Any tips on where else to try would also be welcome.

posted by Paul Slade to Writing & Language (17 answers total)

Can you post an image of the cards in question (I don't think they're included on that twitter thread in the previous post?)?
Even if just a sample, it'll help people decide whether they're likely to actually be able to help you with the full thing, or not.
posted by penguin pie at 3:01 AM on April 13, 2023

Croatian and Serbian are similar languages but Uzbek is totally different, I'm not sure anyone would know if they are able to help you unless they can see the text.
posted by Glier's Goetta at 4:19 AM on April 13, 2023 [1 favorite]

Can you tell what writing system they are, at least? Uzbek was written with Arabic scripts until 1928. Croatian is normally written with the Latin alphabet. Serbian is normally written with the Cyrillic alphabet. Knowing which was used would give a major hint.
posted by McCoy Pauley at 4:45 AM on April 13, 2023 [5 favorites]

Response by poster: Thanks, everyone. I've attached scans of two sample messages in full to my Tweet here. If you prefer to avoid Twitter (and, God knows, I wouldn't blame you), you'll find three full words from the text replacing the mugshot in my Metafilter profile. As you'll see, it's definitely the Latin alphabet.
posted by Paul Slade at 5:07 AM on April 13, 2023

Native speaker here, this is definitely not Serbian or Croatian.
posted by piamater at 5:27 AM on April 13, 2023 [3 favorites]

I'm noticing a few instances of "dh." Albanian isn't the only language that uses those letters together, but it's the only one I'm aware of in that region. Wikipedia seems to agree. Maybe see if you can find an Albanian speaker who can confirm or deny?
posted by nebulawindphone at 5:28 AM on April 13, 2023 [1 favorite]

Another thing to keep in mind is that it might be spelled badly, in an unofficial spelling system, or even in a kind of made-up one. A lot of languages, even if they technically had an official spelling system, lacked real consensus or strong literacy education using it, with the result that lots of people were using a rival system or just winging it. I don't know the history of writing in any of those languages, but this stuff definitely happened in other languages in those parts of the world, and if you're really stuck it's something to consider.
posted by nebulawindphone at 5:48 AM on April 13, 2023 [4 favorites]

I'd wonder if these are coded as well. I'd try to tell if it might be some sort of simple substitution cipher.
posted by gimonca at 6:18 AM on April 13, 2023 [2 favorites]

Are you positive this is another language and not some kind of cipher? Something about the word lengths, sentence structure, distribution of letters and even punctuation (there's a semicolon!) is telling me this is English but encoded for privacy, because, y'know, boarding school.
posted by rpophessagr at 6:19 AM on April 13, 2023 [3 favorites]

Response by poster: You could be right. There's another card in the same batch which definitely use a cipher, and which MeFites decoded for me previously. My assumption that these two are in a foreign language is based on no more than the fact that that's what it looked like to me.
posted by Paul Slade at 6:27 AM on April 13, 2023

Response by poster: I've had a few more thoughts on the cipher idea from Russian/English translator Sandra Mouton over on the Twitter thread. She says:

Could it be written using a crude cipher going like this:
r = l
o = e
d = t
b = p
g = c (or q/k?)
u = i (or a)
z = s
e = o (?)
de dho ruzd = to the last
Yubumozo = Japanese
Brouzo zont u kultom = please send a carton.


The pattern of short, medium and long-ish words reads like English to me (but I could obviously be wrong).

The more I look, the more I'm convinced it's English with the cipher described above. If you keep going, you might be able to come up with a reasonable-looking hypothesis for whole deciphered texts.

Brouzo uzg Tolimk do = Please ask Doring to?
im idz blebol = in its proper?
posted by Paul Slade at 6:49 AM on April 13, 2023 [5 favorites]

Looking at how you've transcribed those.. how would a thick Yorkshire accent of that part of the county and that era be spelled phonetically?
posted by sarahdal at 7:34 AM on April 13, 2023 [1 favorite]

You say in the related Meta-link that these are probably cards from a boy called Turton. In that case, this is definitely a transliterated code based on English words.

DALDEM (bottom left) is TURTON. Above that is ODG = ETC
posted by MinPin at 7:53 AM on April 13, 2023 [2 favorites]

Yes, it's substitution-cipher English. After a bit of fiddling, I transcribe the two cards as:

Toul Shagg. Dhiz Bb iz u zogvor de dho ruzd emo, uz jea virr zoo. I gerrogd uz vorr uz Yubumozo Bbz, dluim enaz, uilzhib enaz, umd nub emoz ef brugoz I gmev umd en imdolozdot im, dhumg enay; pad if jea kod dluim emoz, brouzo te med kod Dagg'z enoz, fel Uamd Gizzj huz zomd no urr dhezo. Huz Tolimk diot ab nj vimohollj dloo jod? Brouzo zomd u kultom lobeld zeem. Flen jealz odg. Ketfloj O Daldem

Toul Shagg. I hebo jea ulo wolj vorr. Dhumg jea wolj nagh fel dho kultom lobeld. Jea gum bigg dho glegaz frevolz dhud ulo ead. Brouzo uzg Tolimk de bad nj fluno im idz blebol peltol em dho proughimk kloom im keet dino fel vhom I geno pugg. Flen jealz odg, Ketfloj O Daldem

Which deciphers as:

Dear Zhucc. This Pp is a secwel to the last one, as you will see. I collect as well as Japanese Pps, train omus, airship omus, ant map ones of places I cnow ant om interested in, thanc omuj; but if you get train ones, please do not get Tucc's omes, for Aunt Cissy has sent me all those. Has Dering tied up my wineherry tree yet? Please sent a garden report soon. From yours etc. Godfrey E Turton

Dear Zhucc. I hope you are very well. Thanc you very much for the garden report. You can picc the crocus flowers that are out. Please asc Dering to put my frame in its proper border on the bleaching green in good time for when I come bacc. From yours etc, Godfrey E Turton

Not perfect, but clear enough what it means...
posted by offog at 8:12 AM on April 13, 2023 [28 favorites]

posted by offog at 8:12 AM on April 13

aka effec or effeq or effek
posted by QuakerMel at 8:18 AM on April 13, 2023 [7 favorites]

dhiz iz yazd de lobeld

I huwo biggot
dho glegaz frevolz
dhud volo im
dho kultom

umt vhigh
jea volo blepuprj
zuwimk
fel vhom jea geno pugg
posted by offog at 8:33 AM on April 13, 2023 [8 favorites]

Response by poster: Thank you, offog - I'm sure you're absolutely right. Godfery was Robin's brother, and both were at Eton at the same time. There's evidence elsewhere in the batch of cards I bought that the family included several keen postcard collectors.
posted by Paul Slade at 9:01 AM on April 13, 2023 [5 favorites]