Gotesch

[Ælfwine]

Gotesch (natively: готəш /ˈɣotəʃ/, hereon referred to primarily as "Gothish," is the last surviving Ostrogothic (East Germanic) language. It is spoken by approximately 4,000 people in the southern part of the Crimean peninsula, largely in the city of Doros (Mangup). History made short, the Ostrogoths reportedly first settled the Crimean peninsula in the 3rd century AD. Likely they became subjects of the Roman Empire, and later the Byzantine Empire in the Principality of Theodoro. In the 8th century, John of Gothia led an unsuccessful revolt against the Khazars. Little is heard of them until Busbecq sought them out and recorded many words in their language. In our timeline, the language probably died out at the tail end of the 18th century, although it is possible it survived as a "Haussprache" until the mid-20th century. For this scenario, I've altered history only so much to keep a handful of Goths alive until the present day.

This conlang is a reboot of my older Goþesch Razde: my attempt to reconstruct what little is known of the Crimean Gothic language. Because a true reconstruction is impossible due to the sparse attestation of Crimean Gothic, I will be filling in the blanks (the actual conlanging part.) Anyway, as is usual today we start with the phonology.

The Gothic phonemic inventory is quite modest, contrasting only 20 consonants. It uses a Cyrillic alphabet inspired by Ukrainian and Russian:

/m n/ м н
/pʰ b tʰ d kʰ (g)/ п б т д к
/f v θ s z ʃ x ɣ/ ф в ѱ с з ш х г
/l r/ л р
/w j/ ў й

Before a velar consonant, /n/ is a velarized [ŋ]. Initially, /b d g/ are partially devoiced: [b̥ d̥ g̊]. /g/ is rare and almost exclusively found in loanwords, while /ɣ/ is the far more common phoneme that represents <г>. In some speakers consonants are allophonically palatalized before front vowels, this is considered nonstandard.

Cyrillic psi (ѱ) was adopted into this language to represent the voiceless dental fricative /θ/ due to its similarity with the biblical gothic letter thiuth – 𐌸. Compare for example Old English and Icelandic adopting runic þ as their letter for /θ/.

Relative to most other Germanic languages, Gotesch has an incredibly simple 6 vowel system:

/i u/ и у
/e ə ɔ/ е э о
/a/ а

In open syllables, vowels are lengthened and diphthongized c.f. ал /aəl/ stone-ACC.S but алс /аls/ stone-NOM.S

And that's it for now. My next post will feature a more detailed list of sound changes. Then we will take a look at declensions, conjugations and adjectives, as well as features like clitic object pronouns, three set of demonstratives for distance (ир, ѱар and єнэр), the very Greek and Turkish influenced number system, amongst other things.

[Travis B.]

How do you explain the lowering of the vowel in the endonym for the language from that in gutiska?

[Ælfwine]

To decide on the sound changes in this conlang, I decided to consult some resources on Crimean Gothic, particularly Stearn's "Crimean Gothic : Analysis and Etymology of the Corpus" as well as a few articles like "The Phonological Systems of Biblical Gothic and Crimean Gothic Compared" by Nielsen. Below is my analysis.

Phonologically it can be stated that Crimean Gothic was a rather conservative language. For example, unlike Biblical Gothic it did not raise the short vowel /e/ to /i/ like in schuuester and reghen (BG swistar and rign). Nor does it appear to have lowered the high vowels before /r/ like in Biblical Gothic. We also see that Crimean Gothic did not turn PGmc /z/ into /r/ as evidenced by fers and wintch. It does appear however to have lowered short /u/ to /o/ before mid/low vowels like in boga (PGmc *bugô). To answer your question Travis, this is why I believe the endonym would be Gotesch instead of Gutesch.

Long /eː/ and /oː/ were raised to /iː/ and /uː/, for example Biblical Gothic mēna > Crimean Gothic mine, unattested BG *stōls > CG stul. The Proto-Germanic diphthongs /ai/ and /au/, which may have been /ɛː/ and /ɔː/ in late BG, resurface as <e> and <oe> in CG, c.f. BG augōna > CG oeghene, possibly resembling a long monophthongal /oː/ sound as it was in Middle Dutch. In some positions the reflex of /ai/ is written <ie>, this happens initially like in iel. In my opinion, Crimean Gothic developed an onglide /j/ in this position before /e/, written here as <ie>.

The inconsistent spelling of vowels in unaccented syllables possibly indicates a reduction to schwa, usually writtten as <e> but sometimes other characters: for example, the stems in the verbs schlipen, breen, geen seem to have <e>, possibly for /ə/, we also see <a> in mycha (c.f. BibGoth mekeis), <o> in ano (BG hana) etc.

In words Busbecq recognized as Germanic we see <h> written initially as in hus and hoef. In words he didn’t recognize as Germanic, but can otherwise be established to be Germanic, we don’t see <h> as in ieltsch or ano. The only exception to be found is lachen, which may have geminated /h/ before /j/ before dropping. (Alternatively, it may have reintroduced this phoneme via analogy.)

One indication that we are dealing with an East Germanic language is that Crimean Gothic had sharpened /jj/ to /d/ as in ada. This kind of development is only found in Biblical Gothic (North Germanic languages like Icelandic had instead turned this phoneme into /g/.)

Occasionally we find devoiced versions of /b/ and /d/ initially, like the <p> in plut or the <t> in tag. In my opinion Crimean Gothic had partially voiced allophones of /b/ and /d/ initially (/g/ was rare in Proto-Germanic). Word finally however we consistently find devoiced plosives and fricatives which signals to me that Crimean Gothic’s final voiced consonants had merged with the voiceless equivalents.

The phoneme /θ/ seems to have been preserved in Crimean Gothic, generally written <tz> like in statz and goltz. Nonetheless it appears that /θ/ had developed to /d/ in some positions, for example, intervocally in bruder, initially before /r/ in tria and in certain words like the article the/tho and possibly thata in malthata/warthata. I think /θ/ had voiced to /ð/ in unstressed syllables and then was fortified to /d/ (and then devoiced again to [t] initially.)

In some words we find /n/ missing finally after /r/, like in kor and baar. In my opinion Crimean Gothic lost /n/ after /r/ similar to Dalecarlian Swedish and Luxembourgish.

In fisct (fisch) and schieten we find <sch> for historical /sk/. We also find <sch> for /s/ before /n/ (sch[n]os), /l/ (schlipen) and /w/ (schuuester), suggesting that /s/ had become /ʃ/ before these consonants as it had in OHG. Despite writing <sch> in ieltsch, borrotsch, and wint[s]ch, I don’t believe Busbeck intended to represent the /ʃ/ sound here, as word finally in Flemish and Dutch it often represented the /s/ sound.

[hwhatting]

I remember reading somewhere that Busbecq was not entirely reliable and that when he recognized a word that was similar to Flemish, he would basically write down the Flemish equivalent. I don't know how true that is, but that could make him an unreliable witness on the question of lowering short vowels.
Are the articles you cite available online? I'd like to file them for future reference.

[Ælfwine]

He was familiar with Middle Dutch/Flemish and Middle High German and probably wrote the words in a way that made sense to him based on similarities to either lamguage. But by examining words in the category that he didn't recognize as Germanic can we definitely come to conclusions. Schuos (for schnos, c.f. Icelandic snör) he didn't recognize as Germanic even though cognates exist, and it seems to have lowered /u/ to /o/ here before a mid vowel. The other possibility is that Crimean Gothic's short /u/ was slightly lowered to begin with, was reproduced by the Greek by an /o/ and therefore writen by Busb. with an <o>. But this doesn't take in account words like brunna "fountain" or sune "sun."

[Travis B.]

I remember reading somewhere that Busbecq was not entirely reliable and that when he recognized a word that was similar to Flemish, he would basically write down the Flemish equivalent. I don't know how true that is, but that could make him an unreliable witness on the question of lowering short vowels.
Are the articles you cite available online? I'd like to file them for future reference.

I have read the same, that a number of Busbecq's transcriptions of words are suspiciously Dutch or German-like due to interference from his knowledge of those languages.

[Ælfwine]

Crimean Gothic Analysis and Etymology of the Corpus: This is by far the best and most comprehensive book on the language. Most of my conclusions come from here. The other article used to be available free on the web but I can no longer find it, but here's the ISBN: 978-3-11-144615-8

The summary of these changes is as follows:

Vn > Ṽː / _{x,xʷ}
e₁ e₂ > eː
jː wː > ɟː gʷː

i u > e o / _{a,oː}

ai > ɛː
au > ɔː
ei eu > iː
iu > yː

eː > iː
oː > uː
ɛː > eː
ɔː > oː
e > i / _…{w,u}

θ > ð / #_-stress
θ > ð / V_V
ð > d
ɟː > dː

sk > ʃ
s > ʃ / _n,l,w
kʷ gʷ hʷ > k g w

x > h
h > x / _j
h > ∅

a e o > ə / -unstressed
i u > ∅ / -unstressed
w,j > ∅ / C_

β > v / V_
β > b / C_

b d g > p t k / #_

This is simply an unfinished rough draft of sound changes. I might change it later on, perhaps even radically. For example, part of me wants to agree with you guys that the vowels remained as they did in Biblical Gothic with no a-mutation either, and I want to develop front rounded vowels from au and eu/iu, though this result isn't quite as clean as the push shift of ɔ: > o: > u: I have going on.

Now time to move on to something arguably a bit more interesting, the grammar!

Nouns and Adjectives:

Several words seem to indicate that Crimean Gothic preserved the masculine nominative singular ending: fers, wint[s]ch, ieltsch, riutsch (In Middle Dutch, final -sch usually represented /s/.) Unlike West or North Germanic, Crimean Gothic sides with East Germanic in preserving this sound as a silibant instead of a rhotic. It also indicates that Crimean Gothic preserved the masculine gender, or merged it with the feminine gender into a common gender like Danish.

An analysis of the adjectives in Crimean Gothic suggests that Crimean Gothic also preserved the neuter gender in words like atochta, wichtgata, and gadelta. The -ata or -ta ending resembles the neuter Biblical Gothic -ata ending.

In Germanic languages that preserve case, greetings like knauen tag are typically used in the accusative. This points to Crimean Gothic having at least two cases: nominative and accusative.

Therefore I propose that Gotesch has three genders, masculine, feminine and neuter and two cases, nominative and accusative. I have considered preserving the genitive, although it had fallen out in neighboring Mariupol Greek in some dialects. However the dative case was lost completely.

[hwhatting]

Crimean Gothic Analysis and Etymology of the Corpus: This is by far the best and most comprehensive book on the language. Most of my conclusions come from here. The other article used to be available free on the web but I can no longer find it, but here's the ISBN: 978-3-11-144615-8

Thanks!