Osroene: A Possible Euphratic IE Language

A forum for all topics related to constructed languages
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Osroene: A Possible Euphratic IE Language

Post by Davush »

I have been toying with the idea of a ‘lost’ IE language (or ‘speculated’ IE’ language) spoken in Mesopotamia for a while now, but none of my attempts so far have been particularly satisfying. In this iteration, I am incorporating some previous ideas/themes, but I’ve also had some new ideas which I would like to develop. This thread will cover history, language, and culture.

Premise: This is an alternative history focusing on what becomes the Osroene Kingdom of Edessa where the language (Osroene) is spoken in the present day. As the title suggests, a ‘possible/speculated’ IE language becomes spoken in roughly the same area as the historical Osroene Kingdom. Historically, Osroene was a Syriac-speaking kingdom in Upper Mesopotamia, situated (precariously) between the Romans to the West and the Parthians to the East from the 2nd century BC to the beginning of the 3rd century AD, Edessa (or Urhay as it was known in Syriac) being its principle city. The main divergence in this alternative history occurs around the 2nd and 3rd century, when Classical Osroene speakers fully enter written history.

Early History and Discovery

Alongside the discovery and decipherment of the Sumerian and Akkadian tablets of Mesopotamia, a number of scattered fragments written in cuneiform in a previously unidentified language (henceforth known as Ancient Euphratic) were also discovered. The earliest of these date from the period corresponding roughly to the Kassite dynasty (1595–1155 BC), and are limited to the southern half of Mesopotamia along the Euphrates, with the largest number being found around Uruk. Initially, little attention was paid to these tablets as the exciting breakthroughs in the decipherment of Akkadian and Sumerian took centre stage.

Curiously, the language is better attested during the Kassite period than Kassite itself, and on the basis of a Euphratic–Babylonian–Kassite lexical list (which is perhaps the most valuable of the Euphratic tablets), a relationship with Kassite was ruled out. This fact has also led to several competing theories, such as that the Ancient Euphratic speakers entered the region along with the Kassites, perhaps then being granted certain privileges, or that they were indigenous, but served a special function during Kassite rule. However, after the Kassite period, the language appears to have vanished from the region, except for a few sporadic instances where groups are mentioned who bear Euphratic-like names. It has been noted that these later mentions come from sources which appear to move northwards up the Euphrates.

The discovery and decipherment of Hittite and other Anatolian languages shortly afterwards led to speculations of possible relationships. Bedřich Hrozný, along with his major breakthroughs in the decipherment of Hittite, noted some striking resemblances between Hittite and this Mesopotamian language, leading him to propose that the two were related, albeit distantly by that point.

During the early twentieth century with the developments in philology and Indo-European studies, a wide range of theories emerged regarding a strange and relatively unknown contemporary language spoken in Upper Mesopotamia, now known as Osroene, ranging from speculations that it was a descendant of Sumerian and forming a group with Hungarian, to it being an obscure branch of Semitic, and everything in between. Nonetheless, after the link to Kassite was ruled out, scholars began to take Hrozný’s proposal more seriously and began examining links between Ancient Euphratic and Neo and Classical Osroene. In recent years, aided by the further discovery of a small number of tablets and a closer analysis of elements such as personal names, there have been more serious proposals that Osroene should be considered (tentatively) within the IE family.

If Osroene is indeed IE, it is thus necessary to account for an earlier stage of PIE before the Anatolian and (tentative) Mesopotamian branches split, which is here called Proto-Anatolo-Mesopotamian.

Timeline of Proto-Anatolo-Mesopotamian

Pre-PIE > Proto-Anatolo-Mesopotamian (PAM) (roughly 4000BC)
> PAM splits into Proto-Anatolian and Proto-Euphratic (presumed distinct and separated by 3000BC).
> Except the early fragments, Ancient Euphratic is then unattested until approximately the 2nd century AD, when Classical Osroene (assuming continuity with Euphratic) becomes widely written in the region.

Next post will show some proposed changes from Proto-Anatolo-Mesopotamian to Ancient Euphratic.
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

The exact chronology of the sound changes from Pre-IE to Ancient Euphratic is unclear, given both the relative scarcity of data and the length of time between its proposed split from Pre-PIE and the earliest attestations. Internal variation and dialectology is also unclear. Additionally, as the earliest attestions come via Akkadian cuneiform (which itself was likely imperfect for Akkadian), the script was likely unable to accurately represent phonological distinctions. However, a number of changes and correspondences have been proposed which appear sound.

Pre-PIE > Ancient Euphratic

A number of sound changes which likely occurred early in the language’s history appear to be shared with Anatolian, namely, the outcome of the stops and laryngeals.

Stops
The outcome of the stops is probably the clearest evidence of an early relationship with Anatolian, as the three-way distinction merges into a lenis/fortis distinction which, by the time of attestation, is wholly determined by stress and vowel length, and therefore a single stop series can be posited for Ancient Euphratic. Word-initially, stops merge into a fortis series most often represented by Akkadian <p ṭ k~q>. The use of Akkadian emphatic <ṭ> for fortis /t/ is unusual, but also supports the idea that they were phonologically unaspirated-unvoiced. The lenis series are most often written with the corresponding signs for Akkadian <b d g>. Although by the time of attestation, further lenition of the lenis series appears to have occurred. Additionally, affrication of *t before *i *j *w appears to have occurred, represented by Akkadian emphatic <ṣ> /ts/. This is partially shared with Anatolian, where it only occurs before *j.

*wód-ṛ > *wādar > wār ‘water’
*ǵrei-t-s > k(i)rēṣ ‘flood’
*gʰew-t-s > qōṣ, kōṣ ‘city walls, city, town’
*sóg-t-s > *sāgiṣ > sāyiṣ ‘belt, buckle’
*dʰéh-ti > *tāti > ṭāṣ ‘to put, to place’
*dʰwégh2-tṛ > *twakχatar > ṣwáqqar ‘noble lady’


Laryngeals
Like Anatolian, Euphratic preserves *h2 and *h3, albeit with some differences in outcomes. In this regard, it also appears more conservative than Anatolian in which they are lost in a number of places. Essentially, both appear to have a fortis/lenis distinction in line with the stops, which results in a merger of them in the lenis positions:

*h2 : Fortis /χ/, Lenis /ʁ/
*h3: Fortis /q/, Lenis /ʁ/

Nonetheless, the exact nature of the realisation of /χ/ and /ʁ/ is unclear given that they are represented by Akkadian <ḥ> and <g>, respectively. Given that the language does not appear to have voiced stops anywhere else, it is assumed that <g> represented a fricative realisation. The development of *h3 to /q/ in fortis positions is surprising, but can be accounted for if we assume that *h3 was /ɣ~ʁ/ or similar, and that Euphratic did not tolerate a voiced-fricative initially, thus merging it with /q/.

Next to stops and *s, *h2 also appears to have merged with /q/, which in some cases may then have become lenited.

*meh2w-ṛ > mēgor /?mēʁor/ ‘time’
*h1esh2-ṛ > *ísχar > *ísqar > ṣqār ‘blood’
*wéth2-ṛ > *wátχar > *wátqar > wáqqar ‘word’ (Hitt: uttar)
*sh2éi-mṇ > *sχēman > sqēman > ṣqēman ‘cord’ (Hitt: išḥiman)
*h2ewh2- > ḥōga /?χoːʁa/ ‘elder, chief’
* plt̥h2-óno-tṛ- > *pəltχānatar > palqānar ‘yoke’
*h2meh1s-h2o- > ḥ(i/a)máṣqa ‘spring’

*h3eḱ-n > ḥwákkan ‘death, plague’
*h3éron > ḥwāran ‘eagle’
*leh3w- > lāgu /?lāʁu/ ‘to fill’
*h3erǵʰi- > ?ḥwāri 'testicle'
Last edited by Davush on 18 Jun 2021 11:06, edited 4 times in total.
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: Osroene: A Possible Euphratic IE Language

Post by WeepingElf »

@Khemehekis - are you sure you posted to the right thread? What do the Reptoids of planet Rathar have to do with Davush's IE conlang?

@Davush - nice stuff so far.
... brought to you by the Weeping Elf
My conlang pages
User avatar
elemtilas
runic
runic
Posts: 3021
Joined: 22 Nov 2014 04:48

Re: Osroene: A Possible Euphratic IE Language

Post by elemtilas »

Davush wrote: 14 Jun 2021 23:54 I have been toying with the idea of a ‘lost’ IE language (or ‘speculated’ IE’ language) spoken in Mesopotamia for a while now, but none of my attempts so far have been particularly satisfying. In this iteration, I am incorporating some previous ideas/themes, but I’ve also had some new ideas which I would like to develop. This thread will cover history, language, and culture.

Premise: This is an alternative history focusing on what becomes the Osroene Kingdom of Edessa where the language (Osroene) is spoken in the present day. As the title suggests, a ‘possible/speculated’ IE language becomes spoken in roughly the same area as the historical Osroene Kingdom. Historically, Osroene was a Syriac-speaking kingdom in Upper Mesopotamia, situated (precariously) between the Romans to the West and the Parthians to the East from the 2nd century BC to the beginning of the 3rd century AD, Edessa (or Urhay as it was known in Syriac) being its principle city. The main divergence in this alternative history occurs around the 2nd and 3rd century, when Classical Osroene speakers fully enter written history.
Be sure to give us the original text of the Letters of Christ and Abgar when you get this sufficiently developed!

I like the premise thus far!
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

elemtilas wrote: 15 Jun 2021 17:18
Be sure to give us the original text of the Letters of Christ and Abgar when you get this sufficiently developed!

I like the premise thus far!
Thanks both! I certainly will.

Some more words found on the Euphratic–Babylonian–Kassite lexical list tablet reveal other possible cognates.

Qorṣwánni-
This appears to be an endo-/exonym describing the Euphratic speakers. A number of etymologies have been proposed, although the most attractive appears to be that it derives from *kʷr-(?s)- ‘cut off’ with the adjectival suffix -wánn-i- meaning ‘coming from’ (compare Hittite -umen- and Luwian -wanni-, IE etymology unclear), thus essentially meaning something similar to ‘the cut off ones’. It is difficult to account for the exact form of the word, as it would be unusual (but perhaps not impossible) for an adjectival suffix to be applied directly to a verbal stem. A participle form in *-(s)h2o- with the -wánni suffix is possible if we assume some reduction has taken place: *kʷr-s-h2ó- ('having been cut off') > Qoriṣqā-wánnis > Qoriṣ(q)wánnis > Qorṣwánnis. If this etymology is indeed true, it sheds light on the status of Euphratic speakers in the region, who were perhaps driven out of their previous territory into Mesopotamia.

Others have further speculated that the Greek ‘Osroene’ (Ὀσροηνή) itself derives from, or was influenced by, a later form of Qorṣwánnis-.

Ṣogwánni-
This is the name of an unknown group/tribe of people who appear to be located distantly. Other have suggested ṣogwánni- is a euphemism for the dead, who are now 'distant'. A proposed etymology attempts to link this to Hittite tūuaz 'from afar', perhaps via *dueh2-uén-i, i.e. 'the ones coming from afar'.

Ṣwáḥḥas
This means 'person', and shares clear similarities with Hittite antuḥsas, of the same meaning, to which it is presumed cognate, although the exact PIE etymology is unclear. Attempts have also been made to connect it to ṣogwánni- as *dueh2-o-s would indeed yield ṣwáḥḥas according to the propose sound changes, but the relationship in meaning between 'person' and the 'distant ones' remains unclear'.

Ṭuṣyánni-
This word seems to mean 'enemy' or 'hostile army'. It has been connected to Hittite tuzzi- 'military camp, army forces', perhaps via an unclear root *dh1-uti- with the addition of a further suffix in Euphratic.

Saqni- 'spring'
This has been connected to Hittite šakkuni- meaning 'spring, well', coming from a proposed root: *sókʷni-.

Verbs
Verbs appear much less frequently in the Euphratic-Babylonian-Kassite lexical list, although they are moderately well attested in some other texts. The following are proposed to have IE etymologies:

iṣ 'to be' < *h1és-ti. The copula appears suffixed to nouns, perhaps under influence influence of or calqueing the Akkadian stative-noun structure. E.g. ṣqās-iṣ 'he is the master/husband'.

ṭāṣ 'to place, put' < *dʰéh1-ti.

íppiṣ 'to take, seize' < *h1ép-ti

qwāṣ 'to kill' < *gʷʰén-ti

wallaṣ 'to pound, to knead dough' < ?*wélh3-ti

From these, it can be seen that the outcome of the 3sg. *-ti ending is regularly -ṣ /ts/ in Euphratic. Some other sound changes can be adduced from these examples.

– Stressed > í, causing gemination of a following consonant, except *we which lowers to wa.
– Resonant + *h3 becomes a geminate resonant, as in *wélh3- > wall-. Some have suggested that this may have phonetically been velarized/pharyngealized /lˤː/ on the basis of evidence from Neo Osroene.
*nti > nṣ, followed by deleted of the /n/ and lengthening of preceding vowel.
Epenthetic i is inserted in stop+stop sequences, *h1ép-ti > ípṣi > íppiṣ.

Participles and Agent Nouns

There are a number of nouns or noun-like words ending in -ṣqās. The similarity to Hittite -šḥas has been proposed, and this is thus suggested to derive from PIE *s-h2ó-(s), which likely had a participle function. As in Anatolian, the common IE passive participles with *tó are absent in Euphratic.

The word olliṣqās 'a shaped loaf' is proposed as coming from *wlh3-s-h2ó-s, i.e. 'that which has been pounded'.

There are also a number of agent noun suffixes, with -(i)(ṣ)ṣes~ṣyas being particularly common. This bears similarity to Luwian/Lycian suffixes which can be derived from *-Vtyo or similar: ollíṣṣyas 'bread-maker' < *wlh3-é-tyos., i.e. 'one who pounds dough'.

Ṣwaḥḥas ollíṣṣes-iṣ /tsʷáχːas olːítsːesits/ 'The person is a bread-maker'
User avatar
KaiTheHomoSapien
greek
greek
Posts: 641
Joined: 15 Feb 2016 06:10
Location: Northern California

Re: Osroene: A Possible Euphratic IE Language

Post by KaiTheHomoSapien »

Ooh I like this a lot. I love the aesthetics of it. It has an almost Lihmelinyan-like feel to it [:D] And I like to see any conlang that uses Anatolian as an influence.
Image
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

KaiTheHomoSapien wrote: 16 Jun 2021 21:23 Ooh I like this a lot. I love the aesthetics of it. It has an almost Lihmelinyan-like feel to it [:D] And I like to see any conlang that uses Anatolian as an influence.
Thanks! I do remember Lihmelinyan being influenced by archaic-IE features, so the resemblance is perhaps expected! [:)]

Notes on the Origins of the Euphratic Speakers
The Euphratic speakers and their obscure history continue to be a source of heated debate, and as of yet, no theory has been conclusively proven. Several points, however, are certain based on the physical evidence:
– They were present in the southern Euphrates region from (at least) around the middle of the 2nd millennium BC.
– Their language is neither Sumerian or Semitic, nor Kassite or Elamite.
– Their language occupied a privileged enough position to have been written.

Of course, these points do not preclude the possibility that Euphratic speakers were present in the region before their attestation in writing; the locations of the tablets are also not necessarily representative of their population centres.

Given the similarities between Euphratic and Anatolian, it is now assumed that the two shared a common ancestor. However, the exact relationship and time/location of divergence between the two branches is far from clear.

Several possibilities have been suggested:

i) Proto-Anatolo-Mesopotamian Hypothesis: In this hypothesis, Anatolian and Mesopotamian form a (briefly) unified clade, splitting from pre-IE and travelling southwards into Anatolia. Given the large divergence by the time of attestation and features unique to Euphratic, the Mesopotamian group is thought to have then split soon afterwards and became isolated from the rest of Anatolian by c. 3000BC. That is to say, in this view Mesopotamian cannot be considered a daughter of Common Anatolian, as evidenced by it not participating in many features that all other (attested) Anatolian languages show.

ii) Early-Split Mesopotamian Hypothesis: This theory states that Mesopotamian never formed a clade with Anatolian, having split significantly earlier. Any shared features between the two are explained as already occurring in ancestral pre-IE. The route and timing of migration into Mesopotamia is unclear. There are arguments for both an Anatolian-route, and a Caucasus-route according to this theory.

iii) Divergent Anatolian Hypothesis: This theory argues for a closer relationship with Anatolian and a later split, in which the apparent divergence of Euphratic is due to it not being considered in the reconstruction of Common Anatolian, with later developments and false etymologies accounting for some issues. The main point used to counter this argument is the outcome of *h2 and *h3 in Euphratic in certain environments not shared with any (attested) Anatolian and some significant lexical differences. Proponents of this argument note that if Common Anatolian is reconstructed with *h2 *h3 according to Euphratic, their loss in certain environments can be considered an innovation of non-Euphratic Anatolian.

Other theories and variations on the above have also been proposed. For now, the exact origins remain unclear. Any opinions on what theory appeals, or any of your own would be welcome! [:D]
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: Osroene: A Possible Euphratic IE Language

Post by WeepingElf »

The "Proto-Anatolo-Mesopotamian Hypothesis" is in some ways similar to the "Southern IE Hypothesis" I am using for (the ongoing revision of) Old Albic. The idea is that there were two primary branches of IE, namely Northern IE, which consists of all extant non-Anatolian IE languages, and Southern IE, which is all extinct, with only the Anatolian languages preserved in writing. Southern IE would have been spoken, besides Anatolia, all along the Danube and in Western Europe by the Bell Beaker people, before it was eclipsed by Northern IE languages, in the Early Bronze Age in the Danube area and in the Late Bronze Age in Western Europe. The last holdout of Southern IE besides Anatolia may have been pre-Celtic Britain, and I am re-creating this language as Old Albic. Currently, I am working on its ancestor, Proto-Hesperic, the language of the Bell Beaker people. (The sketches of Old Albic and Proto-Hesperic that are currently on FrathWiki are out of date, since they were built on an earlier model of the origin of the Elves which had assumed a much earlier divergence from IE proper but turned out to be untenable.)
... brought to you by the Weeping Elf
My conlang pages
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

WeepingElf wrote: 17 Jun 2021 17:12 The "Proto-Anatolo-Mesopotamian Hypothesis" is in some ways similar to the "Southern IE Hypothesis" I am using for (the ongoing revision of) Old Albic. The idea is that there were two primary branches of IE, namely Northern IE, which consists of all extant non-Anatolian IE languages, and Southern IE, which is all extinct, with only the Anatolian languages preserved in writing. Southern IE would have been spoken, besides Anatolia, all along the Danube and in Western Europe by the Bell Beaker people, before it was eclipsed by Northern IE languages, in the Early Bronze Age in the Danube area and in the Late Bronze Age in Western Europe. The last holdout of Southern IE besides Anatolia may have been pre-Celtic Britain, and I am re-creating this language as Old Albic. Currently, I am working on its ancestor, Proto-Hesperic, the language of the Bell Beaker people. (The sketches of Old Albic and Proto-Hesperic that are currently on FrathWiki are out of date, since they were built on an earlier model of the origin of the Elves which had assumed a much earlier divergence from IE proper but turned out to be untenable.)
Interesting! Nice to know others are working on similar ideas. I'd be interested in seeing how Proto-Hesperic evolves from "Southern IE".

Where did the Laryngeals Go?

(Apologies for the piecemeal presentation, but I am horrible at keeping track of notes, and posting them helps organise my thoughts a bit better.)

The discovery of Euphratic had shed some new light on the nature of PIEs laryngeals. The outcomes can be summarized thus:

*h2 > /χ(ː)~ʁ/, /q/
*h3 > /χʷ~ʁʷ~o/, /qʷ~qo/

Word-initially, *h2 is well represented as <ḥ>, which was likely uvular /χ/.

*h2ewh2- > <ḥōgas> /’?χoːʁas/
*h2meh1-s-h2o- > <ḥamáṣqas> /’?χamátsqas/

After long vowels, *h2 appears to lenite, being spelled with Akkadian <g>, as in <ḥōga> above. In some cases, however, *h2 gains a plosive realization as represented by Akkadian <q~k>, particularly when following another stop or *s,i.e. the sequence *sh2 seems to regularly yield <ṣq> /?tsq/: e.g., <ṣqār> ‘blood’ has been proposed to derive from h1ésh2-r (Hitt: išḥar) with some reanalysis of the root. Participle-like forms also show <-ṣqā->, which has been proposed as deriving from *s-h2ó-.

It is thus unclear whether *h2 was thus originally a stop /q(ː)/ in Proto-Anatolo-Mesopotamian (as Kloekhorst proposes), which became fricated in certain environments, or vice-versa, a fricative /χ(ː)/ which fortified.

Most importantly, the outcome of word-initial *h3é is <ḥwá~ḥwā> /χʷa~χʷaː/, although few examples exist. This is unlike Hittite, which appears to have merged it with *h2é /χa/ in the same words.

*h3érōn- > <ḥwāran> /?ˈχʷaːran/ ‘eagle’ (Hitt: ḥaran-)
*h3éḱn- > <ḥwákkan> /?ˈχʷakkan/ ‘death, plague’
*h3éwi- > <ḥwāwis> /?ˈχʷaːwis/ ‘sheep’ (Hitt: ḥawi-)

The Euphratic evidence thus suggests *h3 was labialized in a stage ancestral to it and Anatolian, which was subsequently lost in Hittite. Inter- and post-vocalically, as well as in clusters, the outcome of *h3 in Euphratic is less clear due to paucity of examples and possible merges/loss in this position.

Of the existing examples, the pair of verbs <táḥḥoṣ> 's/he takes' and <ṣqwāṣ> 'they take' are especially intriguing. Initially, it was unclear whether the two represented the same stem as <ṣqwa-> could be the reflex of many pro-forms. However, it has recently been proposed that the two derive from PIE *deh3- ‘to take, give’, where the reflex of *h3 is clear in Euphratic but lost in Anatolian (cf. Hittite: dāi~tāi, danzi~tanzi). The following has been proposed:

PIE: *déh3-ti > Proto-AM: *Taxʷ-ti > Euphratic: táχχots (where unstressed *CʷC regularly > CoC)
PIE: *dh3-énti > Proto-AM: *Txʷ-énti > Euphratic: tsqʷaːts (via *tχʷ > tsqʷ, and regular *wénC > wāC).

Although the above strongly points towards *h3 being labialised in Proto-AM, it is nonetheless still unclear whether *h3 should be reconstructed as voiced. The fortis/voiceless reflex in <taḥḥoṣ> can be explained as either continuing an unvoiced segment, or due to devoicing caused by the following unvoiced stop, and in <ṣqwāṣ> as part of the regular sound change in which all voiced stops became unvoiced initially.
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: Osroene: A Possible Euphratic IE Language

Post by WeepingElf »

Davush wrote: 18 Jun 2021 10:40
WeepingElf wrote: 17 Jun 2021 17:12 The "Proto-Anatolo-Mesopotamian Hypothesis" is in some ways similar to the "Southern IE Hypothesis" I am using for (the ongoing revision of) Old Albic. The idea is that there were two primary branches of IE, namely Northern IE, which consists of all extant non-Anatolian IE languages, and Southern IE, which is all extinct, with only the Anatolian languages preserved in writing. Southern IE would have been spoken, besides Anatolia, all along the Danube and in Western Europe by the Bell Beaker people, before it was eclipsed by Northern IE languages, in the Early Bronze Age in the Danube area and in the Late Bronze Age in Western Europe. The last holdout of Southern IE besides Anatolia may have been pre-Celtic Britain, and I am re-creating this language as Old Albic. Currently, I am working on its ancestor, Proto-Hesperic, the language of the Bell Beaker people. (The sketches of Old Albic and Proto-Hesperic that are currently on FrathWiki are out of date, since they were built on an earlier model of the origin of the Elves which had assumed a much earlier divergence from IE proper but turned out to be untenable.)
Interesting! Nice to know others are working on similar ideas. I'd be interested in seeing how Proto-Hesperic evolves from "Southern IE".
It is all still very much under construction. The final Proto-Hesperic will look quite different from what is currently on FrathWiki, so much is sure. There will be more vowels, a more complex accent rule, and very different inflectional paradigms. But nothing is in dry cloths yet, so it would be premature to present it.
... brought to you by the Weeping Elf
My conlang pages
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Osroene: A Possible Euphratic IE Language

Post by Salmoneus »

It's been too long since I was last doing a basal PIE language for me to remember enough about PIE to comment meaningfully.

One thing, though: you talk about 'Pre-PIE', but your parent language looks a lot like PIE. Do you have any differences in mind? Indeed, it looks a lot like LPIE - I assume you're assuming Proto-Indo-Hittite wasn't that different from Late PIE?


[surprised not to see mention of the Gutians, btw!]
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

Salmoneus wrote: 18 Jun 2021 14:07 It's been too long since I was last doing a basal PIE language for me to remember enough about PIE to comment meaningfully.

One thing, though: you talk about 'Pre-PIE', but your parent language looks a lot like PIE. Do you have any differences in mind? Indeed, it looks a lot like LPIE - I assume you're assuming Proto-Indo-Hittite wasn't that different from Late PIE?


[surprised not to see mention of the Gutians, btw!]
I haven't given a great amount of thought as to the differences between Pre-IE and Late PIE, mostly because I am not knowledgeable enough and PIE scholarship is a rabbit hole (any recommendations on papers/sources discussing pre/early IE would be welcome!), so I am assuming PIH was not too different from Late PIE. An obvious difference would be how the stops are reconstructed, with pre-PIE here possibly using Kloekhorst's suggested /tː ʔt t/ series, but since they all collapse into a single series in Euphratic anyway (although I may rethink this), I've just used the conventionally reconstructed values. There will probably also be some grammatical/morphological features retained reflecting earlier PIE.

On the topic of Anatolian stops, I am somewhat perplexed by the idea that the primary distinction was primarily fortis/lenis, even word initially, with the fortis series being geminate. (This may just be me misunderstanding the material though). I know some scholars propose the lenis series were voiced initially (and intervocalically?), but to my mind, this would seem more intuitive to understand as allophonic gemination of unvoiced stops after stressed syllables? E.g. (ignoring vowels)

*ta > /ta/ (?/tta/)
*da dha /da/ (?/ta/)

*ata > /átta/
*ada adha > /aːda/ (?/aːta/).

Any thoughts/speculations on this and/or early IE would be interesting to hear!

As for the Gutians, I do actually have them in my notes connecting them with a possible early appearance of the Euphratic speakers! (In my previous attempt, the Gutian theory was the main one, but this connection is less well-established here).
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: Osroene: A Possible Euphratic IE Language

Post by WeepingElf »

I think the correct name for the common ancestor of Late PIE (the common ancestor of the non-Anatolian IE languages) and Anatolian is Early PIE. At least, that's what I understand that term to mean. And reconstructing its morphology is not easy since we are dealing with just two descendant entities and no reliable outgroup - so how do you tell which of the two innovated where?

That said, quite something can be reconstructed. The phonology of Early PIE does not seem to be much different from that of Late PIE, as the phonology of Anatolian can be reduced to Late PIE phonology with little difficulty. What is likely is that the phonetic values of some of the phonemes may have changed from Early to Late PIE - but the overall structure of the phonology seems not to have changed much (which also tells us that the difference in time depth is not great). Regarding morphology, those items that Late PIE and Anatolian share can be quite safely reconstructed for Early PIE, and it seems as if Anatolian is more archaic in this regard - there's quite something in Late PIE, such as the feminine gender and the tripartite verb aspect system, that is absent from Anatolian, but not really much in Anatolian that is absent from Late PIE. This means that one can get a good idea of Early PIE, at least good enough for developing conlangs from it, by projecting Anatolian morphology, minus some idiosyncrasies that look like Anatolian innovations such as the ergative case in -anza, back onto the Late PIE phonology.
... brought to you by the Weeping Elf
My conlang pages
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

WeepingElf wrote: 18 Jun 2021 15:56 I think the correct name for the common ancestor of Late PIE (the common ancestor of the non-Anatolian IE languages) and Anatolian is Early PIE
[...]
Thanks - yes that's pretty much also my understanding of the situation, hence using the "conventional" reconstructions.

Mitanni, (Indo)-Iranic, and the Gutians: A Euphratic Connection?

The single attestation of a peculiar word in a Euphratic text relating to horse training has led to a number of theories. This word is <aswállaḥ>, likely meaning ‘horse trainer’. The form of the word is particularly curious, as the usual word for horse in Euphratic is <íkkos>, fairly well attested, which is the expected outcome of PIE *h1éḱu-o-, so the presence of the root <aswa-> in <aswállaḥ> cannot be derived from the Euphratic reflex, but is clearly similar to the (Indo)-Iranic reflexes. Interestingly, Mitanni shows the word <a.aš.šu.uš.ša.an.ni>, which has been interpreted as ‘āśva-san-ni’, and thus <aswállaḥ> has been proposed as deriving from a Mitanni or Mitanni-like term, where the suffix has been replaced by the more Euphratic <-llaḥ> seen in several other agent nouns.

Although the history of the Mitanni is itself murky, this has led to several interesting speculations. One is that the Euphratic speakers migrated southwards, perhaps being driven out by the Mitanni themselves during this time of upheavel. This theory is attractive in many ways: the Mitanni are thought to have arose sometime around the 16-17th millennium BC in northern Mesopotamia after a power vacuum was created following the Hittite destruction of the Semitic-speaking Yamhad dynasty. It is thus possible that the Euphratic speakers attempted to control the region, but were overwhelmed by the Mitanni, fleeing to southern Mesopotamia.

Another theory holds that <aswa-> is not a necessarily Mitanni loan, but could represent a loan from any early (Indo)-Iranic-like language. In this theory, the Euphratic speakers may have been present in the Zagros mountains alongside early Iranic speakers and the Kassites, perhaps entering Mesopotamia as allies of the Kassites.

A Gutian connection has also been highlighted, in which scenario the Euphratic speakers may either be the continuation of the Gutians themselves or a related group. The merit of this argument is that the Gutians appear to have been centred around the southern Mesopotamian plains after their conquest, which corresponds to the area where Euphratic is attested. Nonetheless, if a significant Gutian population did remain, it is curious that the only significant piece of evidence is a reference from the Old Babylonian period to a non-Akkadian speaking group in Adab.

A Plural Morpheme in <-aḥ>

The suffix <-aḥ> appears somewhat frequently in Euphratic and has been proposed as a plural or collective marker, deriving from IE *-eh2. It is mostly attested in words referring to animals and ethnic/geographic/linguistic groups. The main attestations with secure meanings and/or IE etymologies are (sg. forms in brackets):

aswállaḥ ‘horse trainers’ (? Mitanni loan)
íkkaḥ (íkkos) 'horse' (< *h1éḱu-eh2) (-oḥ would be expected, perhaps reflecting generalised -aḥ?)
qwánnaḥ (qwánnas) ‘woman/women’ (< *gʷén(h2)-eh2)
ḥwitraḥ (ḥwītar) ‘wild beasts’ (< ? *h2uéid-r-eh2)
qūwaḥ (qūwas) ‘cattle’ (< *gʷōu-eh2)
Qorṣwánnaḥ (Qorṣwánnis) ‘the Qorṣwanni’
Elāmaḥ (Elāmos) ‘the Elamites

aswállaḥ (together with íkkaḥ) notably occurs only in the following sentence of a tablet:

aswállaḥ íkkagam in kisállom ḥannistarís tāgi ?/aswállaχ íkkaʁam in kisállom χannistarís táːʁi/
horse.trainer-PL horse-PL.ACC in(to) court-ACC king-GEN take-3sg.PST
The horse trainers took the horses to the king's court.

It is interesting to note that nouns in -aḥ appear to usually take singular agreement. <kisállom> is clearly an Akkadian loan from <kisállum> of the same meaning. It has further been speculated that the unusual Akkadian prepositions <ina> and <ana> may be borrowings from/influenced by Euphratic <in> and <na>.

Having established some possible theories regarding the Euphratic speakers, and that the language is most likely IE, the next posts will hopefully look at how Euphratic may relate to Classical and Neo Osroene.
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

A Closer Look at Euphratic Plosives

The outcome of PIE plosives in Euphratic requires further attention. Although the spelling system is defective and thus any conclusions are tentative, the evidence is compelling in some aspects. The outcome of the stops is also the main reason suggesting Euphratic's early split from Anatolian, as it differs in several respects which cannot be post-Anatolian developments.

Essentially, the outcome of word-initial *T D Dʰ in Euphratic is > T=D, but remains (partially) distinct. Medially, the outcome is split between T=D, and D=Dʰ. Within IE, the *T=D merger with distinct *Dʰ is unique to Euphratic, not being found in any other IE branch. This also offers some clues to the nature of pre-Anatolian PIE stops, discussed below.

Word Initially:
*T and *D merge.
*Dʰ shows a different outcome in several instances, particularly so for *ǵʰ *gʷʰ, and for *dʰi *dʰu. Initial *p *bʰ appear to merge in words with secure etymologies.

Word Medially:
*T = *D in *éD, where they also geminate.
*D = *Dʰ in *óD and other non-, and lengthen the previous syllable.


Detailed examples of the outcomes Euphratic words with a secure IE etymology:
Spoiler:
The T=D=!Dʰ outcome is most clearly illustrated by the following sets:

T: *ḱēr- kēr > ‘heart’
D: *ǵenu- kinnu > ‘knee’
Dʰ: *’ǵʰim-ro ímmara > ‘field’

T: *kʷer- qwār- > ‘to cut’
D: *gʷén(h2)- qwann- > ‘woman’
Dʰ: *gʷʰén- χwann- > ‘to kill’

Thus it would appear that at least for non-plain velars, Euphratic has: *K=G, whereas *Gʰ shows a different outcome.

It is more difficult to ascertain a parallel outcome for coronals and plain velars due to not enough secure attestations to make a full comparison. However, *dʰ appears to behave differently from *t *d before *u~w:

D: *dui-jo-no- > tuyanállaḥ ‘second in rank’
Dʰ: *dʰúh2-i-mo > ṣuχχimas ‘smoke’
BUT: *dʰeh1-ti > tāṣ ‘he says’

Additionally, a different outcome for *dʰy~*dʰi has been proposed if the PIE 2sg. imperative ending *-dʰi is reconstructed for a small number of irregular imperatives ending in -s, rather than -ṣ, which is usual outcome for *ty *dy, e.g. *tiéh-no > ṣyānas ‘autumn’; *dyēw-t-s > ṣyōṣ ‘day’.

The labial series all appear to merge as /p/ initially.

Medially: The outcome appears to be split between T=D, and D=Dʰ, preceding causes T=D, and causes D=Dʰ with lengthening (as in Anatolian). The *D series undergo a further stage of elision in Euphratic, sometimes eliding completely.

Medial *D=Dʰ before *ó: *wód-ṛ: waˀdar > wādar > wār. *yugom: yuˀgam > yūgan > yūʁan.

Medial *T=D before é: *négʷ-mo-nt-: > náqqoman

Distinct *Dʰ before *é: *dʰégʰ-m: tēgam > ṣīyan.
Evidence from Akkadian Loans
Spoiler:
In general, Euphratic is faithful to the common Akkadian spellings of loanwords, although some discrepancies in spelling should be considered. Notably, the earlier texts show confusion of Akkadian initial voiced stops and voiced-geminates with their voiceless counterparts, hinting that Euphratic speakers did equate these with their own voiceless consonants and thus did not have an initial/geminate voiced series. E.g. Akkadian <ina libbi> ‘inside, within’ is seen as <in lippi>. Later texts are more faithful to the Akkadian, perhaps indicating greater acculturation by that stage and/or the gain of a true voicing distinction under Akkadian influence.
Proto-Anatolo-Mesopotamian Plosive System
The merger of initial the unvoiced–voiced (T=D) series appears to be unique to Euphratic, occurring in no other branch. Although on first sight this is a puzzling development, a phonological argument can be used to explain this.

Namely:
*T=D merger results from a shared +[fortis] feature.
*Dʰ is marked as –[fortis], thus remaining distinct in some outcomes, particularly as –[fortis], or +[lenis] are more phonologically prone to lenition.
*T D sharing +[fortis] can be explained if *T is assumed to be unvoiced and tenuis or possibly lightly aspirated, and *D as originally having some glottalic feature, e.g., +constricted glottis, later lost and reinterpreted as +[fortis], thus merging with *T.
– Medially after *ó, *D appears to lose its +[fortis] feature. Perhaps indicating a different word-medial allophone in PAM.

The following has been proposed for the Proto-Anatolo-Mesopotamian stop-system:

/t(ʰ) tˀ~ˀd d̥~dʰ/

This analysis is essentially a compromise between the conventional and glottalic theories, in that it emphasizes that *D series were likely not just plain-voiced but had another glottalic feature (exact nature unknown), and that the *Dʰ series likely already had some sort of +[lenis] feature by the time of Proto-Anatolian.

(Any comments/criticisms/questions welcome!)
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: Osroene: A Possible Euphratic IE Language

Post by WeepingElf »

Fine. My personal view of the Early PIE stop system no longer involves any glottalized stops, though. I have found a rather simple solution to the question of the root structure constraints: the *T set was aspirated, while the *D and *Dh sets were as in the traditional reconstruction. This puts *T and *Dh in one class(both have the feature [+breath]) at the exclusion of *D. I even have a hunch that the *D set once were especially weak sounds, perhaps even voiced spirants or such, which may explain the near-absence of *b: it merged with *w, thereby also explaining the *wl- and *wr- initial clusters. There are, after all, no glottalized stops or anything like that in Uralic, which probably is the closest living kin of IE, nor in the other "Mitian" families (the Itelmen ejectives are clearly an innovation of that language). So the PIE system may have started with something Uralic-like with a set of voiceless stops and a matching set of voiced spirants. Then, the spirants hardened to voiced stops, excepting the labial one which became *w, and the old stops acquired aspiration. Either at this time or before that, a dissimilation rule removed one of the two spirants/voiced stops in a root where there were two of them, Then, in some morphemes, perhaps caused by some lost vowel feature, the aspirated stops were voiced, but this resulted in breathy-voiced stops instead of merging with the old voiced stops.

It is uncertain when the aspirated stops lost aspiration. The aspiration of the *T set in Germanic and Armenian may have been an archaism, conserved by the devoicing of the *D set, but it may just as well have been an innovation of these two branches (independently, of course). The developments in Italic, Greek and Indo-Aryan of course require an unaspirated *T set, as these branches innovated voiceless aspirates (in Italic ahd Greek, from the *Dh set, in Indo-Aryan, from *TH clusters) which remained separate from *T. The remaining languages, where *Dh merged with *D and *T remained distinct, do not really say anything decisive, but in most (all?) of them, *T is not aspirated. (Some people claim that *T was aspirated in Proto-Celtic, but that seems to be a minority opinion.)

As for a *T/*D merger, I have once found a claim that this happened in Phrygian, but others say that this language had an Armenian-like shift, and yet others that it had merged *Dh with *D. Otherwise, I haven't heard of any IE language where this happened (disregarding Tocharian, where all three merged, except that *t and *d have different outcomes).
... brought to you by the Weeping Elf
My conlang pages
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

WeepingElf wrote: 21 Jun 2021 09:39 Fine. My personal view of the Early PIE stop system no longer involves any glottalized stops, though. I have found a rather simple solution to the question of the root structure constraints: the *T set was aspirated, while the *D and *Dh sets were as in the traditional reconstruction.
[...]
Thanks - that is all very interesting. I don't hold strongly to any particular theory of PIE stops in actual reconstructions myself. Euphratic is intended to be a clear "odd one out" within PIE, which is why I like the *T=D idea, even though it's not attested in other branches. Your theory of aspirated *T with usual *D Dʰ would also work for Euphratic if it's assumed that there's a constraint against word-initial voicing (as assumed for Anatolian). With this theory, *T would have lost its aspiration early enough, but *Dʰ retained it (not sure how likely this would be, but it doesn't strike me as totally implausible), so the +[breath] of *Dʰ puts it in opposition to initial *T *D due to loss of voicing on *D and aspiration on *T. Medially, it's a bit stranger. The *D=Dʰ outcomes could be explained by the voicing opposition persisting a bit longer medially, although *éT=éD is still a bit of a mystery (but perhaps not totally bizarre as vowel qualitiy is attested as affecting things like this, I think)...
User avatar
WeepingElf
greek
greek
Posts: 531
Joined: 23 Feb 2016 18:42
Location: Braunschweig, Germany
Contact:

Re: Osroene: A Possible Euphratic IE Language

Post by WeepingElf »

Davush wrote: 21 Jun 2021 10:17
WeepingElf wrote: 21 Jun 2021 09:39 Fine. My personal view of the Early PIE stop system no longer involves any glottalized stops, though. I have found a rather simple solution to the question of the root structure constraints: the *T set was aspirated, while the *D and *Dh sets were as in the traditional reconstruction.
[...]
Thanks - that is all very interesting. I don't hold strongly to any particular theory of PIE stops in actual reconstructions myself.
Nor do I! The "*T was aspirated" theory is just an idea of mine which seems plausible to me, but I don't know what kind of problems there may be with it which I don't see for lack of knowledge. And "Mitian" is of course a highly speculative matter. I feel that the similarities between those languages (IE, Uralic, Turkic, Mongolic, Tungusic, Yukaghir, Chukotko-Kamchatkan, Eskimo-Aleut) are better explained by descent from a common ancestor than by anything else, but there probably is no way proving that because there are too few cognates to establish regular sound correspondences. At least, DNA studies seem to suggest that a human population fanned out across Northern Eurasia from somewhere near Lake Baykal at the end of the last ice age, and it would seem plausible that those people spoke Proto-Mitian, but we all know that genes and languages do not always travel together!
Euphratic is intended to be a clear "odd one out" within PIE, which is why I like the *T=D idea, even though it's not attested in other branches.
Of course. It is not in principle illegitimate to do something in an IE conlang that no IE natlang has done. My own Old Albic, for instance, is a full-fledged active-stative language of a like I have never seen in any IE natlang. The idea came to me when I read suggestions that an early stage of PIE may have been active-stative, and I decided to develop that idea into a conlang.
Your theory of aspirated *T with usual *D Dʰ would also work for Euphratic if it's assumed that there's a constraint against word-initial voicing (as assumed for Anatolian). With this theory, *T would have lost its aspiration early enough, but *Dʰ retained it (not sure how likely this would be, but it doesn't strike me as totally implausible),
My hypothesis would at least require that for Italic, Greek and Indo-Aryan, which may be the biggest problem with it. One would expect *Dh also to lose breath and merge with *D - which is exactly what happened in most branches! Hence, I think that the *Th-D-Dh system was unstable and the gap at *T (i.e., voiceless unaspirated) was closed by deaspirating *Th, probably already in Late PIE. *Dh was not deaspriated in some languages because there was no *D gap. For Hesperic, BTW, I assume that the *T gap was remedied by devoicing *D instead, leading to a *Th-T-Dh system (i.e., the same as in Armenian, where there are good reasons to assume that the voiced stops of Old Armenian were actually breathy-voiced).
so the +[breath] of *Dʰ puts it in opposition to initial *T *D due to loss of voicing on *D and aspiration on *T. Medially, it's a bit stranger. The *D=Dʰ outcomes could be explained by the voicing opposition persisting a bit longer medially, although *éT=éD is still a bit of a mystery (but perhaps not totally bizarre as vowel qualitiy is attested as affecting things like this, I think)...
Well, in a *T-D-Dh system, *D can swing both ways as it shared non-aspiration with *T and voicedness with *Dh, so you are IMHO fine.
... brought to you by the Weeping Elf
My conlang pages
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

WeepingElf wrote: 21 Jun 2021 17:05 ested as affecting things like this, I think)...
[...]
Well, in a *T-D-Dh system, *D can swing both ways as it shared non-aspiration with *T and voicedness with *Dh, so you are IMHO fine.
Thanks! All very useful information. I look forward to reading about Old Albic (when it is ready, of course).

The Zezkānni and Ṣaṣqānū: Euphratic Speakers?

Although the mysterious identity of the Euphratic speakers has led to many theories and speculations, two small pieces of evidence in particular appear promising.

In a cuneiform tablet describing the reign of Hittite Hattusili I (c. 1650–1620 BC), we find mention of a people called Zezkānni (ze.ez.ka.an.niš). Here, the Zezkānni are described as hostile towards Hattusili, and as neighbours of the Isuwa. In the tablet, during one of Hattusili I’s many campaigns against neighbouring groups, he destroys the Zezkānni lands and drives them out, causing them to flee southwards, or, “to Ḥanigalbat, the land of the King of the Hurrians”, which is interpreted as referring to the newly emerging Mitanni-Hurrian power to the south. No other mention of the Zezkānni has been discovered in Hittite writings, and very little also of the Isuwa until it becomes the centre of the Luwian state centuries later. Nonetheless, if the Zezkānni where geographically close to the Isuwa, this would place them somewhere in the upper reaches of the Euphrates at the periphery of Anatolia proper.

Several centuries later, and further south, we then find brief mention of several individuals who are described as Ṣaṣqānū bearing Euphratic-sounding names in a recently discovered tablet attributed to 'the son of Kurigalzu' (Kassite King, Kurigalzu I, died c. 1375 BC). Here, land-grants and official positions are awarded to these Ṣaṣqānū individuals in Nippur due to their loyalty to the Kassites; they are also described as being “brothers to my ancestors”. It is around this time that tablets written in Euphratic begin to appear in this area.

Although the evidence is fragmentary, a link between the Zezkānni and the Ṣaṣqānū appears promising, with the dates and locations of the tablets matching a southward migration. Thus, the Euphratic speakers have been proposed as likely candidates for the identity of the Zezkānni/Ṣaṣqānū, although this is obviously a tentative conclusion.

Additionally, several etymologies have been proposed on the assumption that Ṣaṣqānū is a word of Euphratic origin; the presence of <-ṣqā-> in particular points to this fact, as it is a common participle-like form seen in many Euphratic words. Ṣīyas appears to have been a major Euphratic god, and is proposed as deriving from PIE *dyēws, thus cognate with Zeus. One etymology suggests that Ṣaṣqānū is a reduced/corrupted form of ṣīya-ṣqā-(nū) from PIE *dyēw-s-h2ó-(nt-), i.e., ‘the people of Ṣīyas’. Another suggestion is that it is instead an Akkadian exonym, originating in a mocking term for the Euphratic language due to its many <ṣ> and <q> sounds, akin to ‘barbarian’.

Assuming Zezkānni = Ṣaṣqānu = Euphratic speakers, a brief timeline could be:

c. 4000BC: Proto-Anatolo-Mesopotamian splits from PIE.
c. 3000BC: Proto-Euphratic and Proto-Anatolian have split from each other. Likely already present in Anatolia.
c. 2000BC: Early Euphratic speakers occupy (south-)eastern Anatolia, perhaps pushed eastwards by other Anatolian groups?
c. 16000BC: Zezkānni located at periphery of eastern Anatolian before being driven southwards.
c. 1375BC: Ṣaṣqānū mentioned near Nippur. Some names appear to be Euphratic.
c. 1375-1100: Euphratic is moderately well attested in writing around the mid-southern Euphrates.
c. 1100+BC: Euphratic disappears. Sporadic reference to groups who may be identified with the Ṣaṣqānū.
Davush
greek
greek
Posts: 670
Joined: 10 Jan 2015 14:10

Re: Osroene: A Possible Euphratic IE Language

Post by Davush »

Medieval Osroene

As the period around 1200 –1400AD is where I want to focus on in this alternative history, I should probably present an overview of Medieval Osroene, spoken around this point. I won’t go into great detail on the specific history in this point; it is enough to note that Classical Osroene had flourished between c. 600 – 1100AD, where it is fully attested in vocalized scripts. Old Osroene before it, appeared c. 300 BC, mostly in unvocalized Aramaic. As is to be expected, much of the phonology and various grammatical features of Osroene mirror the surrounding languages of the time, namely, Aramaic and its various dialects, as well as Greek.

Initially believed to be a divergent Anatolian language, a lost Semitic language, or an isolate, a link to Euphratic is becoming increasingly likely as scholars continue to make breakthroughs. Nonetheless, the time distance of nearly a millennium from the last attestations of Euphratic, Osroene’s relative isolation amid much larger non-IE languages, and the little that can be firmly said about Euphratic naturally make it difficult to ascertain a firm relationship.

If Osroene is indeed IE, it would be unique in many respects.

Phonology
(This is just an overview for now. More detailed posts later…)

Vowels
/i u/
/e ə o/
/a ɑ~ɔ/

/ɑ~ɔ/ is closer to [ɑ~ɒ] in closed and unstressed syllables, and tends to be closer to [ɔː] in open stressed syllables.

/i e/ and /u o/ merge as [ɪ ʊ] in closed syllables and word finally, usually spelled with /e o/. This gives a closed syllable/word final inventory of /ɪ ʊ *ɛ a ɑ~ɔ ə/.

/ə/ is a full vowel, appearing commonly in stressed closed syllables. It has several allophones depending on surrounding consonants, ranging between [ɨ~ə~ʌ].

/*ɛ/ may also exist as a marginal phoneme.


Consonants

(*) indicates marginal phonemes, or phonemes mostly found in loan words.

/p t k *c q/
/b d g *ɟ/
/f θ ð χ ħ *ʁ~ʕ/
/s *z ʃ/
/m n/
/r l/
/β j/

The lenition of singleton consonants /p t k~q/ > /f θ χ/ appears to have occurred at roughly the same time as in Syriac/Aramaic, perhaps under additional Greek influence. /β/ likely also results from a shift of /w > β/ during the Classical period. There appears to be some amount of confusion with /ʁ~ʕ/.

/c ɟ/ are allophones of /k g/ before front vowels, although they can be considered marginally phonemic before /ə/.

Stress and Other Details
Stress is mostly fixed on the penultimate in Medieval Osroene, although the innovation of some word-final stress can also be observed. This change appears to have occurred sometime during the mid-late Classical period. The vowels /ɔ i u/ are lengthened in open, stressed syllables.

Stressed /ə/ causes following consonants to geminate.
Word intial CC clusters are common, and may be realized with an initial epenthetic /ə/, particularly if a previous word would result in a CCC cluster.

Nouns: Plurals

Medieval Osroene has several classes of noun which form their plurals in different ways. The most common involve the following vowel alternations of the final vowel:

Singular – Plural
a – ɔ: qɔːna – qɔːnɔ ‘woman’; ħiːða – ħəðrɔ ‘animal’
a, ɔ, u – e: əmra – əmre ‘field’; sɔːħa – sɔːħe ‘person’, cənnu – cənne ‘knee’
a, ɔ – ɔne: ħnəʃta – ħnəʃtɔne ‘king’; paqa – paqɔːne ‘guardian’
nɔ – n: ʃrɔːnɔ – ʃɔːrɔn ‘cloud

There are several common irregular plurals, including:
ˈənnɔ – ˈejɔn ‘path, way’

Genitive /m/
Genitives are formed by the use of the Genitive particle /m/, which functions similar to the Romance <de>. Old Osroene and early Classical Osroene appear to preserve a distinct genitive case, although this has almost entirely disappeared by the medieval period. Phonetically, this is realized /(ə)m/ or /mə/.

ˈəmra m ˈqɔːna 'the woman’s field'
ˈpaqa mə ˈħnəʃta 'the king’s guardian'
ˈejɔn əm ˈsɔːħe 'the people’s paths'
ˈħəðrɔ m ˈəmre 'the animals of the fields'

Orthography
As Greek influence grew in the region, Osroene eventually switched (by c. 600AD) to the Greek alphabet due to its ease of representing vowels and greater suitability compared to the Syriac script. Osroene developed its own variant in order to adapt the script to various phonemes not found in Greek, similar to the development of the Coptic script. Thus, the form presented here is more of a "stand in" than an actual representation.

π τ κ κχ /p t k q/
β δ γ /b d g/
φ θ χ η (γ) /f θ χ ħ (ʁ~ʕ)/
σ ξ /s ʃ/
μ ν ρ λ ου ι /m n r l w~β j/

ι ου /i u/
ε υ ο /e ə o/
α ω /a ɑ/

Notably: eta <η> is used for /ħ/. Osroene probably developed its own letter for /ħ/, à la Coptic, but <η> will function as a stand-in (especially considering it derives from ħeth.
Similarly, xi <ξ> used as a stand in for /ʃ/. And the digraph <κχ> for /q/. I could use archaic qoppa and san for /q/ and /ʃ/, but these are a pain to type.

An example (including above genitive phrases) in the script:
υμρα μ κχωνα
πακχα μ ηνυξτα
ειων μ σωηε
ηυδρω μ υμρε


κχωνω σω νσυκκχω τω ηνυξτα νσεσα ν υμρα μ ηυδρω
/qɔnɔ sɔ nsəqqɔ tɔ ħnəʃta nsesa n'əmra m'ħəðrɔ/

Any orthographical suggestions welcome.
Post Reply