Oscan Reconstruction

Discussion of natural languages, or language in general.
User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Oscan Reconstruction

Post by kanejam »

Hello everyone!

So Znex (if they're still keen) and I had the idea to make a modern day Oscan descendent, but the first problem we have is that there isn't a whole lot of inscriptions in Oscan. The morphology, in particular the verbal morphology, and the vocabulary are both lacking. So this thread is a request for help in reconstructing Oscan.

Most of what I have is from reading through Buck's Grammar of Oscan and Umbian. Umbrian is also an interesting language with some neat innovations, but for now we'll focus on Oscan. Buck wrote that 'Oscan is the Gothic of the Italic dialects. In the conservatism and transparency of its vowel-system, it is rivalled only by Greek of all the Indo-European languages. Diphthongs are preserved intact in all positions.' Oscan also underwent anaptyxis between voiceless stops and liquids or nasals: /aragɛtu:d/ for Latin argentō - 'with silver'.

Let's start with the phonology.
/m n/
/p t k/
/b d g/
/f s x/
/w l r j/

Monophthongs in stressed syllables: /iː u uː e eː ɛ ɛː a aː o/
Monophthongs in unstressed syllables: /i u e ɛ a o/
/ai ɛi oi au ou (ɛu)/

- Consonants can appear geminate.
- Stress is always word-initial
- The fricatives /f s/ were voiced intervocalically. As far as I can tell, /x/ never occurred intervocalically. /x/ might also be realised as [h]. Intevocalic /s/ never rhotacised, giving the 1st declension genitive plural ending /a:su:m/ rather than Latin -ārum.
- There is no rounded velar. It merged with /p/ making it P-Italic rather than Q-Italic :wink: . /eu/ only appears in loans from Greek.It is unclear whether /i/ is a separate phoneme or just an orthographic quirk. /o:/ only exists word finally and comes from final /a:/, basically making it an allophone although it is distinguished orthographically. /ɛː/ is rare.

Unresolved questions:
- Were both /f/ and /x/ voiced intervicalically? In Oscan, unlike Latin, they remain separate from the voiced stops and so they appear between vowels. Was /x/ actually /x/ and if so, was it voiced? I don't know how common /x/ is between vowels but it could be /h/ instead of /x/.
- What is the exact value of /e e:/? Proto-Italic *i and *ē are both written with the same symbol in the native alphabet and seemed to be differentiated only through quantity. However in Latin alphabet inscriptions, they are both written <i> rather than <e>, so maybe they were closer? Would /ɪ ɪː/ be likely? This sort of ties in with the next question:
- Is /i/ a separate phoneme? I'm thinking not as it only appears in certain conditions or where it's likely to be /i:/. This is a bit complex to try to explain.
- Was stress initial? It was definitely so in Proto-Italic and early Oscan, but may have changed to a weight based system like Latin. I don't know if there's a whole lot of evidence to argue either way but it might have come about throug language contact. The trouble with that is that Oscan had more contact with the other Sabellic languages and Greek colonies.


This post is getting long, so here is a little bonus section: Orthography

/m n/ m n
/p t k/ p t k
/b d g/ b d g
/f s x/ f s h
/w l r j/ v l r i

/(i) iː u uː e eː ɛ ɛː a aː o/ i ií u uu í íí e ee a aa ú
/ai ɛi oi au ou (ɛu)/ aí eí úí av úv ev

- z was also used to indicate /ts/. x was unused.
- Long consonants were written doubled.
- Long vowels were often marked but not completely consistently, especially in inflectional endings. The genitive ending /a:su:m/ that I mentioned earlier was written -asúm.


Any comments or thoughts on the unresolved questions? I would love some feedback or help with this. If people are interested then I will carry on with nominal morphology next.
Last edited by kanejam on Thu Jul 10, 2014 5:28 pm, edited 3 times in total.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
Terra
Avisaru
Avisaru
Posts: 571
Joined: Tue May 24, 2005 10:01 am

Re: Oscan Reconstruction

Post by Terra »

I'm interested. Morphology, please!

Also, does Oscan continue any roots that died out in Latin?

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

Terra wrote:I'm interested. Morphology, please!

Also, does Oscan continue any roots that died out in Latin?
Yay! I'll give a general overview of morphology and then probably focus on nouns. And yes it does have many words not found in Latin, even with the limited number of inscriptions.
Here are a few, some of which are also found in Umbrian:
- her- meaning 'want, wish', cognate to English yearn, almost completely displacing *wel, Latin volō, velle
- túvtú meaning 'people, translating best with Latin civitas
- egmú meaning 'thing', where Latin would use 'rēs'. If Oscan had take over half of Europe instead of Latin, we might say Egmotoutic instead of Republic (from egmú túvtikú, my idea of a loan translation of rēs pūblica)
- eítuvú meaning money, Latin pecunia
- íním for 'and', Latin et.
These are just a handful, there are many more.

Here are more characteristics of Oscan, from a slightly more diachronic viewpoint.
Phonology
Shared with Umbrian:
- /*kʷ *gʷ/ > /p b/
- extensive syncope of non-initial short vowels e.g. actud compared with Latin agitō
- /*nd *ks/ > /nn ss/
- retention of /s/ before liquids and nasals where it is lost in Latin
- retention of /a/ in medial syllables where it is weakened to e or i in Latin.
- retention of medial /f/ lost in Latin
- final *ā > ō
- /*kt *pt/ > /xt ft/; the Oscan for Octavius is Úhtavis, with accompanying syncope of the *o in the nominative ending *-os.
- most /ns/ > /f/ except in inflectional endings where it changed to /ss/. In Umbrian it changed to f in all positions. This one is a bit funny and the conditions aren't completely clear for Oscan.

Only in Oscan:
- the anaptyxis mentioned earlier
- the shifting of monophthongs, including /*ō/ merging with /*ū/ into /u:/.
- intervocalic /s/ is preserved as [z] and final /d/ is preserved everywhere.

Morphology
The nouns fall into the same five declensions of nouns as Latin. The main difference is that the consonant stems are still distinct from the i-stems. The fifth declension has only the tiniest number of words and the fourth is pretty poorly attested as well. Oscan has the same three genders as Latin, masculine, feminine and neuter. For those who aren't familiar with Latin, here is a rough overview. I will go into more detail in another post.
- 1st declension, a-stems, mostly feminine
- 2nd declension, o-stems, mostly masculine and neuter. The neuter words have slightly different forms: both the singular nominative and accusative resemble the masculine accusative and both the plural nominative and accusative resemble the a-stem singular nominative. The short o in the accusative ending is syncoped, obscuring some words: Húrz is Hortus. There's also a subset of io-stems.
- 3rd declension, split into consonants and i-stems. Mixture of all three genders
- 4th declension, u-stems
- 5th declension, i-stems

The cases are the same as Latin except that the singular Locative remains distinct. The cases are: nominative, accusative, vocative, genitive, dative, ablative, locative. Nouns can be either singular or plural.

The verbs are also very similar to Latin, although there are many holes in the paradigm. Almost all attested forms are in the third person. The main differences are a distinction between primary and weakened secondary forms to change tense, whereas Latin generalised the primary forms and used tense suffixes. Oscan also has some unique ways of marking the perfect which aren't found in Latin or Umbrian.

Syntax is almost identical, and as Buck says, 'there are no uses of mood and tense which cannot be paralleled in Latin.' The genitive and locative are used in some places where Latin would be more likely to use prepositions.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
dhok
Avisaru
Avisaru
Posts: 859
Joined: Wed Oct 24, 2007 7:39 pm
Location: The Eastern Establishment

Re: Oscan Reconstruction

Post by dhok »

Are you quite sure Oscan had no nasals? Seems unlikely.

User avatar
Kereb
Avisaru
Avisaru
Posts: 463
Joined: Sat Feb 12, 2005 12:59 pm
Location: Flavor Country™
Contact:

Re: Oscan Reconstruction

Post by Kereb »

smartass.
<Anaxandridas> How many artists do you know get paid?
<Anaxandridas> Seriously, name five.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

dhok wrote:Are you quite sure Oscan had no nasals? Seems unlikely.
Yup! It shared this feature with Umbrian and Latin.

Thought I forgot something :oops: fixed now
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
Sleinad Flar
Lebom
Lebom
Posts: 124
Joined: Tue Feb 27, 2007 11:18 pm
Location: Coriovallum, Germania Inferior

Re: Oscan Reconstruction

Post by Sleinad Flar »

kanejam wrote:Unresolved questions:
- Were both /f/ and /x/ voiced intervicalically? In Oscan, unlike Latin, they remain separate from the voiced stops and so they appear between vowels. I don't know how common /x/ is between vowels but it could be /h/ instead of /x/.
- What is the exact value of /e e:/? Proto-Italic *i and *ē are both written with the same symbol in the native alphabet and seemed to be differentiated only through quantity. However in Latin alphabet inscriptions, they are both written <i> rather than <e>, so maybe they were closer? Would /ɪ ɪː/ be likely? This sort of ties in with the next question:
- Is /i/ a separate phoneme? I'm thinking not as it only appears in certain conditions or where it's likely to be /i:/. This is a bit complex to try to explain.
- Was stress initial? It was definitely so in Proto-Italic and early Oscan, but may have changed to a weight based system like Latin. I don't know if there's a whole lot of evidence to argue either way but it might have come about throug language contact. The trouble with that is that Oscan had more contact with the other Sabellic languages and Greek colonies.
Answers from Wallace, The Sabellic Languages of Ancient Italy, 2007:

- Yes, /f/ was voiced intervocally, just like /s/; it was occasionaly transcribed with <β> in Greek script. However <h> indicates a glottal fricative /h/ in initial positions (not /x/) and a hiatus between vowels, so no voicing there.
- Wallace mentions 'three pairs of palatal vowel phonemes' (i.e. /ę, e, i/ and their long counterparts). Examples: pedum, ídík, leginum, teerum, fíísnam, viíbius. The examples are shown here in rising hight, so yeah, /e/ <í> would have been higher than /ę/<e> (I don't get his notation either, ę stands for lowered e, i.e. for ε?). But note that /e/ is written as <í> and /o/ as <ú>, because the Etrucscan alphabet only had 4 vowel symbols (a,e,i,u); it doesn't automatically mean that /e/ and /o/ were higher than usual. The Oscan-Greek transcriptions use <η> and <ο>, as expected.
- Wallace mentions the possibility that vowel length in general was only maintained in word-initial and radical syllables; outside of these double spellings are rare. Having said that, it's peculiar he uses leginum as an example of /i/, while all other examples are in initial syllables. You might be on to something here.
- In view of this and other evidence (syncopation of short vowels in medial open syllables), word-initial stress is suggested for all Sabellic languages, also in historical times. However, nothing is known for certain.

EDIT:
The genitive ending /a:su:m/ that I mentioned earlier was written -asúm.
Note that in the light of Latin -ārum (< PIt *-āsom), and aforementioned vowel shortening in non-intial syllables, this ending probably simply was /asom/.
"Was ist ist, was nicht ist ist möglich"
http://sleinadflar.deviantart.com

User avatar
Particles the Greek
Lebom
Lebom
Posts: 181
Joined: Tue Sep 17, 2013 1:48 am
Location: Between clauses

Re: Oscan Reconstruction

Post by Particles the Greek »

dhok wrote:Are you quite sure Oscan had no nasals? Seems unlikely.
The only known picture of a speaker of Oscan shows someone with no nose, so it may be due to a racial or cultural peculiarity.
Non fidendus est crocodilus quis posteriorem dentem acerbum conquetur.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

Thanks for the reply! It's good to have some feedback.
- Yes, /f/ was voiced intervocally, just like /s/; it was occasionaly transcribed with <β> in Greek script. However <h> indicates a glottal fricative /h/ in initial positions (not /x/) and a hiatus between vowels, so no voicing there.
Okay I'm happy with f being voiced intervocalically. What evidence does Wallace give for h being /h/? I thought it was more likely that it was /x/ because of clusters like ht in rehtas. Although it would explain where the h in stahínt comes from which always confused me a bit. However I'm not convinced that the expected *míheí would be /mi.ei/. I'll have another look at attested words with intervocalic h, they're definitely not many. Maybe in words such as rehtas it simply indicates lengthening of the e caused by the shift /x/ > /h/.
The Oscan-Greek transcriptions use <η> and <ο>, as expected.
Yes but the Oscan-Latin transcriptions use <i> consistently where Oscan writes í which makes me unsure about its value. There isn't any reason I've transcribed ú as /o/, it could just as well be /O/.
- Wallace mentions the possibility that vowel length in general was only maintained in word-initial and radical syllables; outside of these double spellings are rare. Having said that, it's peculiar he uses leginum as an example of /i/, while all other examples are in initial syllables. You might be on to something here.
I can't find the example word leginum so I can't comment, but maybe it came from *ī ? Anyway, reading over Buck, the only instances of i not coming from *ī are before a /j/ which seems to be just orthographic, and from samprasana (vocalisation of a semivowel) of /j/ such as in jo-stems after syncope of of the o. But if you're correct that long vowels are only distinguished word-initially then this can simply merge with the reflex of *ī without a problem.

Double spellings do occur in non-initial syllables but maybe these represent either vowels in hiatus or archaic spellings. Since I'm trying to reconstruct this for the purpose of making a descendent rather than a faithful representation of the ancient Oscan language, shall we just agree that stress is word-initial and long vowels only contrast in this position? For the moment though double consonants can appear anywhere but word-initially (and only ss word finally).
Note that in the light of Latin -ārum (< PIt *-āsom), and aforementioned vowel shortening in non-intial syllables, this ending probably simply was /asom/.
I'm not sure why I thought it was *-āsōm; maybe I was misled by all the ú's that Buck argues represent /u:/ such as in tanginúd.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

As always, if anyone wants to jump in and help they would be welcome.

Case Endings

I'll give four columns, the first will be Oscan, then an example word, then the Latin, then the Proto-Italic. Forms in green are not direct descendents of the Proto-Italic forms. Forms in italics are my own assumption.


First Declension

These continue the -eh2 nouns of PIE such as tewtéh2 (seem familiar?). After Sleinad Flar's post I've decided to leave all vowels short.

Sg
Nom... -ú ... túvtú ... -a ... *-ā
Voc ... -a ... túvta ... -a ... *-a
Gen ... -as ... túvtas ... -ae ... *-ās
Dat ... -aí ... túvtaí ... -ae ... *-āi
Acc ... -am ... túvtam ... -am ... *-ām
Abl ... -ad ... túvtad ... -ā ... *-ād
Loc ... -aí ... túvtaí ... (-ae) ... *-āi
Pl
Nom/Voc ... -as ... túvtas ... -ae ... *-ās
Gen ... -asúm ... túvtasúm ... -ārum ... *āsom
Dat/Abl ... -aís ... túvtaís ... -īs ... *-ais
Acc ... -ass ... túvtass ... -ās ... *-ans

There's very little of interest here. The main points are that Oscan retains the -ās of Proto-Italic whereas Latin -ae is innovated from somewhere. Also of interest is the Proto-Italic ablative ending in -ād which was generalised from the thematic nouns. I'm unsure of the vocative forms though; Buck seems to imply as fact that the ending in P-It. was short but I can't find evidence supporting this.

Second declension is a bit bigger so that will come shortly!
Last edited by kanejam on Sat Jan 18, 2014 11:12 pm, edited 2 times in total.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
Sleinad Flar
Lebom
Lebom
Posts: 124
Joined: Tue Feb 27, 2007 11:18 pm
Location: Coriovallum, Germania Inferior

Re: Oscan Reconstruction

Post by Sleinad Flar »

You're welcome!
kanejam wrote:Okay I'm happy with f being voiced intervocalically. What evidence does Wallace give for h being /h/? I thought it was more likely that it was /x/ because of clusters like ht in rehtas. Although it would explain where the h in stahínt comes from which always confused me a bit. However I'm not convinced that the expected *míheí would be /mi.ei/. I'll have another look at attested words with intervocalic h, they're definitely not many. Maybe in words such as rehtas it simply indicates lengthening of the e caused by the shift /x/ > /h/.
Are you sure this rehtas isn't Umbrian? I cannot find it in the Wörterbuch des Oskisch-Umbrischen, although it does mention U. rehte. In Umbrian <h> was used in this position to indicate vowel length.
The evidence for h being /h/ seems to be that it's transcribed as <h> in the Latin alphabet as well, e.g. heriiad, herest.
Yes but the Oscan-Latin transcriptions use <i> consistently where Oscan writes í which makes me unsure about its value. There isn't any reason I've transcribed ú as /o/, it could just as well be /O/.
It could also be /U/. Unfortunately, Wallace is very concise on this matter. A table of comparison between the different writing systems would have been nice.
I can't find the example word leginum so I can't comment, but maybe it came from *ī ? Anyway, reading over Buck, the only instances of i not coming from *ī are before a /j/ which seems to be just orthographic, and from samprasana (vocalisation of a semivowel) of /j/ such as in jo-stems after syncope of of the o. But if you're correct that long vowels are only distinguished word-initially then this can simply merge with the reflex of *ī without a problem.
Leginum appears in Wörterbuch as an acc. sg., cognate to La. legionem, but exact meaning unknown (gathering, troop). Historically, this was a stem in *-iH-on-,*-iH-n-. Latin generalized the strong stem, Osco-Umbrian the weak stem: *-iH-n- > *-ī-n- > -in-, so yes, you are correct.
(This also might explain why i doesn't appear in initial syllables, or at least is very rare: in Italic (also in Celtic) *-iH- was shortened to *-i- in word-initial pretonic syllables, e.g. *wiHrós > *wiros)
Double spellings do occur in non-initial syllables but maybe these represent either vowels in hiatus or archaic spellings. Since I'm trying to reconstruct this for the purpose of making a descendent rather than a faithful representation of the ancient Oscan language, shall we just agree that stress is word-initial and long vowels only contrast in this position? For the moment though double consonants can appear anywhere but word-initially (and only ss word finally).
I'm perfectly happy with that.
(BTW double spellings are also used for glides, e.g. heriiad /he.ri.jad/ or something similar.)
I'm not sure why I thought it was *-āsōm; maybe I was misled by all the ú's that Buck argues represent /u:/ such as in tanginúd.
You're right, according the Buck (p. 88) -úd, -ús, -úm (with alternative spelling -ud, -us, -um) represent /-ūd, -ūs, -ūm/. According to Wallace however, these were shortened to /-ud,-us,-um/ (which neatly fits his model). In the light of the other endings, I was wrong assuming -úm represented /-om/; Wallace's /-um/ or even Buck's /-ūm/ is way more probable.
"Was ist ist, was nicht ist ist möglich"
http://sleinadflar.deviantart.com

zompist
Boardlord
Boardlord
Posts: 3368
Joined: Thu Sep 12, 2002 8:26 pm
Location: In the den
Contact:

Re: Oscan Reconstruction

Post by zompist »

kanejam wrote: Voc ... -a ... túvta ... -a ... *-a
You mark this as a reconstruction, but why would it be different from the nominative? Latin only has a vocative in the II singular declension... is there any evidence that Oscan or Umbrian had more?

(Great thread, by the way. The not-quite-Latinness of the other Italic languages is fascinating.)

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

zompist wrote:
kanejam wrote: Voc ... -a ... túvta ... -a ... *-a
You mark this as a reconstruction, but why would it be different from the nominative? Latin only has a vocative in the II singular declension... is there any evidence that Oscan or Umbrian had more?

(Great thread, by the way. The not-quite-Latinness of the other Italic languages is fascinating.)
Thanks! This is taken from Buck. There are no examples of an Oscan vocative, but around fifty Umbrian examples. In the Umbrian, they are always written with a, rather than varying between a and the more common u. Buck takes this as an example that it is continuing the 'old vocative'. Looking at the PIE forms, I can't see where this 'old vocative' might have come from. Maybe the Oscan should just be .
Sleinad Flar wrote:You're welcome!
Are you sure this rehtas isn't Umbrian? I cannot find it in the Wörterbuch des Oskisch-Umbrischen, although it does mention U. rehte. In Umbrian <h> was used in this position to indicate vowel length.
The evidence for h being /h/ seems to be that it's transcribed as <h> in the Latin alphabet as well, e.g. heriiad, herest.
I can't find rehtas again, so maybe it was from Umbrian. Here are a list of words that I could find with non-initial h: ehtrad, saahtúm, Úhtavius, feíhuss, kahad, Verehasíúí and also the prefix eh. The h's in the first three are from *kt; the double aa in saahtúm is from 'secondary lengthening' although it doesn't say what that is. Verehasíúí is particularly important because it comes from *werhasjo-, meaning the anaptyxis that is characteristic of Oscan occurs even between *r and *h. The fact that it was transcribed with h in the Latin script doesn't lend much weight as I'd imagine that /x/ would also be transcribed h. The only other spelling that might be suitable would be ch. I'm not sure if there are enough Greek inscriptions with /h/ in them but I'll have a look.

Diachronically, Buck talks of *gh becoming h, with no mention of what happens intervocalically. When describing the phonology, he just transcribes h as h, and again doesn't mention how it might be pronounced non-initially.
Sleinad Flar wrote:
Yes but the Oscan-Latin transcriptions use <i> consistently where Oscan writes í which makes me unsure about its value. There isn't any reason I've transcribed ú as /o/, it could just as well be /O/.
It could also be /U/. Unfortunately, Wallace is very concise on this matter. A table of comparison between the different writing systems would have been nice.
It's unlikely that ú was ever /U/ because its Latin script counterpart is always o. The fact that í was written in Latin as i worries me a little bit; but going by my sense of aesthetic, I'll just decide that e is /E/ and í is /e/.
Sleinad Flar wrote:Leginum appears in Wörterbuch as an acc. sg., cognate to La. legionem, but exact meaning unknown (gathering, troop). Historically, this was a stem in *-iH-on-,*-iH-n-. Latin generalized the strong stem, Osco-Umbrian the weak stem: *-iH-n- > *-ī-n- > -in-, so yes, you are correct.
(This also might explain why i doesn't appear in initial syllables, or at least is very rare: in Italic (also in Celtic) *-iH- was shortened to *-i- in word-initial pretonic syllables, e.g. *wiHrós > *wiros)
But this isn't a problem as *wiros will just end up as vír. I'm happy leaving short /i/ out completely.
Sleinad Flar wrote:
Double spellings do occur in non-initial syllables but maybe these represent either vowels in hiatus or archaic spellings. Since I'm trying to reconstruct this for the purpose of making a descendent rather than a faithful representation of the ancient Oscan language, shall we just agree that stress is word-initial and long vowels only contrast in this position? For the moment though double consonants can appear anywhere but word-initially (and only ss word finally).
I'm perfectly happy with that.
(BTW double spellings are also used for glides, e.g. heriiad /he.ri.jad/ or something similar.)
Yes, I was thinking more about the double ii stems of the second declension found in names, but for now we can probably ignore them or assume that ii in that case represented /iji/.
Sleinad Flar wrote:
I'm not sure why I thought it was *-āsōm; maybe I was misled by all the ú's that Buck argues represent /u:/ such as in tanginúd.
You're right, according the Buck (p. 88) -úd, -ús, -úm (with alternative spelling -ud, -us, -um) represent /-ūd, -ūs, -ūm/. According to Wallace however, these were shortened to /-ud,-us,-um/ (which neatly fits his model). In the light of the other endings, I was wrong assuming -úm represented /-om/; Wallace's /-um/ or even Buck's /-ūm/ is way more probable.
Upon rereading Buck, he does think that the ending was *-āsōm. I don't think -úm always represents /um/, as in the 2nd declension acc sg, which comes from *-om. At one point however Buck mentions that maybe a change of o > u occurred before final m but then at another point says that, unlike Latin, short *o was preserved before final *m.

Is your Wallace book available online? I might have to have a look at it. What about any other books with a good view of Proto-Italic? In the meantime I'll get around to writing up the second declension.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

I just found that a final s before another word starting with s is written h. I'm not sure what that says about it but maybe final s is quite weakly produced and that's how it contrasts with final ss.

Second Declension

Sg
Nom ... -s ... húrz ... -us ... *-os
Voc ... -e ... húrte ... -e ... *-e
Gen ... -eís ... húrteís ... -ī ... *-oi?
Dat ... -úí ... húrtúí ... -ō ... *-ōi
Acc ... -úm ... húrtúm ... -um ... *-om
Abl ... -ud ... húrtúd* ... -ō ... *-ōd
Loc ... -eí ... húrteí ... -ī ... *-ei
Pl
Nom/Voc ... -us ... húrtús* ... ... *-ōs
Gen ... -um ... húrtúm* ... -ōrum ... *-ōm
Dat/Abl ... -úís ... húrtúís ... -īs ... *-ois
Acc ... -úss ... húrtúss ... -ōs ... *-ons

The s of the nominative singular disappears after an l or r. If this causes a weird consonant cluster then an anaptyctic vowel appears, as in Latin. An example is *agros > ager (attested in Umbrian). Another note is that the asterisked forms should all have u rather than ú but are much more common with the latter spelling. To avoid confusion I will use u from here on.

The singular genitive is taken from the consonant stems, but otherwise the Oscan continues the Pr-It forms nicely. The Latin plural genitive was taken from the athematic stems. Latin's nominative plural comes from something like *-oi which supposedly occurred in Greek as well. These are for masculine nouns (and whatever rare feminine nouns there are, if they exist). The neuter words have slightly different forms for the nominative and accusative cases:
Sg
Nom/Acc ... -úm ... prúftúm ... -um ... *-om
Pl
Nom/Acc ... -ú ... prúftú ... -a ... *-ā

This sets the scene for all neuter nouns: no distinction between nominative and accusative and the nominative singular are the same as the accusative singular and the plural are the same as the 1st declension nominative singular. Otherwise the endings are the same as the rest of their declension. If that was confusing then I'm basically saying that neuter nouns behave exactly the same as in Latin.

There is a subset of second declension nouns known as the io-stems because the original Pr-It stems end in *-jo rather than just *-o. The /j/ undergoes samparasarana in some cases but remains consonantal elsewhere (unlike Latin):

Pl
Nom ... -is ... degetasis ... -ius ... *-jos
Voc ... -ie ... degetasie ... -ie ... *-je
Gen ... -ieís ... degetasieís ... -iī (sometimes -ī) ... *-joi?
Dat ... -iúí ... degetasiúí ... -iō ... *-jōi
Acc ... -im ... degetasim ... -ium ... *-jom
Abl ... -iud ... degetasiúd ... -iō ... *-jōd
Loc ... -ieí ... degetasieí ... -iī ... *-jei
Pl
Nom/Voc ... -ius ... degetasiús ... -iī ... *-jōs
Gen ... -ium ... degetasiúm* ... -iōrum ... *-jōm
Dat/Abl ... -iúís ... degetasiúís ... -iīs ... *-jois
Acc ... -iúss ... degetasiúss ... -iōs ... *-jons

Neuter Nom/Acc
Sg ... -im ... memnim ... -ium ... *-jom
Pl ... -ú ... memniú ... -ia ... *-jā

Edit: well this is embarrassing... tangins is a consonant stem :P
Last edited by kanejam on Sun Jan 19, 2014 3:58 am, edited 1 time in total.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
Sleinad Flar
Lebom
Lebom
Posts: 124
Joined: Tue Feb 27, 2007 11:18 pm
Location: Coriovallum, Germania Inferior

Re: Oscan Reconstruction

Post by Sleinad Flar »

I'll respond to your comments later (I want to read Buck a little more before I respond), but for this:
Is your Wallace book available online? I might have to have a look at it. What about any other books with a good view of Proto-Italic? In the meantime I'll get around to writing up the second declension.
Only in Uzbekistan (http://uz-translations.net/?category=an ... ient_italy).
As for proto-Italic, I only have two etymological dictionaries: Michiel De Vaan - Etymological dictionary of Latin and the other Italic languages and Jürgen Untermann - Wörterbuch des Oskisch-Umbrischen. There's some more literature on the historical developments of Latin, e.g. Gerhard Meiser - Historische Laut- und Formenlehre der lateinischen Sprache and Michael Weiss - Outline of the Historical and Comparative Grammar of Latin, which (of course) treat proto-Italic is an intermediate between PIE and Latin, and have a few excursi on Osco-Umbrian.
"Was ist ist, was nicht ist ist möglich"
http://sleinadflar.deviantart.com

zompist
Boardlord
Boardlord
Posts: 3368
Joined: Thu Sep 12, 2002 8:26 pm
Location: In the den
Contact:

Re: Oscan Reconstruction

Post by zompist »

Excursus is 4th declension... the Latin plural is excursūs. There's an amusing bit on this in Geoff Pullum's Great Eskimo Vocabulary Hoax.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

@Sleinad Flar, I've found a few more examples of h marking hiatus between vowels instead of an actual consonant. I'm not sure if I'll be able to make a decision about h without some definitive evidence either way.
zompist wrote:Excursus is 4th declension... the Latin plural is excursūs. There's an amusing bit on this in Geoff Pullum's Great Eskimo Vocabulary Hoax.
Wait where did excursus come from?
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

Gentile

There is a special subset of io-stems which are basically iio-stems. In Latin and Oscan, the gentile is a patronymic formed by infixing /j/ into the ending of the father's (I'm assuming) praenomen. An example from Latin is Tullius from Tullus. However, in Oscan, io-stems are very common as praenomina, and the gentiles from these end up with two i's. Some examples of Oscan names are Vibis Smintiis, Dekis Rahiis, Stenis Kalaviis. It appears to be pronounced /ji/ in this case because it causes the consonant doubling, and in the oblique cases it has it's usual pronunciation of /ij/.


Third Declension

The third declension is much less coherent in Oscan than Latin, so I will just start with the i-stems and then tackle the mighty consonant stems later. Third declension nouns are of all three genders. There are a few examples of i-stems becoming io-stems, so maybe if I were to collapse the case system a bit then that would be one of the first changes I make.

SG
Nom ... -s ... aídil ... -s ... *-is
Voc ... -s ... aídil ... -s ... *-is?
Gen ... -eís ... aídileís ... -is ... *-eis
Dat ... -eí ... aídileí ... -ī ... *-ei
Acc ... -im ... aídilim ... -em ... *-im
Abl ... -id ... aídilid ... -ī ... *-īd
Loc ... -eí ... aídileí ... -ī ... *-ēi
PL
Nom/Voc ... -ís/is ... aídilis ... -ēs ... *-ejes
Gen ... -ium ... aídilium ... ium ... *-jōm
Dat/Abl ... -íss ... aídilíss ... ibus ... *-ifos
Acc ... -íss ... aídilíss .. īs ... *-ins

The are no examples of neuter i-stems in Oscan. In the plural they most likely ended in -iú, but it's unclear what the ending in the singular would be. However Umbrian follows Latin's lead, either turning the final *i into an e or deleting, as in the Latin sedīle and animal.

Of interest are the Nom Pl, which has the two alternate forms. The first is the natural continuation of the Pr-It, but the second is from *-īs by analogy with -ās and -ōs of the first two declensions. Also, there are no examples of an Acc Pl, although if the Pr-It form *-ins is continued then it will merge with the Dat/Abl form. Lastly, íss of the Dat/Abl is completely cognate with Latin -ibus, even though they don't look similar at all.

Next up, consonant stems!
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
Salmoneus
Sanno
Sanno
Posts: 3197
Joined: Thu Jan 15, 2004 5:00 pm
Location: One of the dark places of the world

Re: Oscan Reconstruction

Post by Salmoneus »

kanejam wrote: Some examples of Oscan names are Vibis Smintiis, Dekis Rahiis, Stenis Kalaviis.
Shock discovery: Oscans actually all eastern european mafia.
Blog: [url]http://vacuouswastrel.wordpress.com/[/url]

But the river tripped on her by and by, lapping
as though her heart was brook: Why, why, why! Weh, O weh
I'se so silly to be flowing but I no canna stay!

User avatar
Particles the Greek
Lebom
Lebom
Posts: 181
Joined: Tue Sep 17, 2013 1:48 am
Location: Between clauses

Re: Oscan Reconstruction

Post by Particles the Greek »

The problem I have most with Oscan is trying to adjust to the use of <i> and <u> to represent mid vowels.
Non fidendus est crocodilus quis posteriorem dentem acerbum conquetur.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

araceli wrote:The problem I have most with Oscan is trying to adjust to the use of <i> and <u> to represent mid vowels.
I don't find it so much a problem; remember that it is only a transcription of the native alphabet. I got used to it pretty quickly and now I like it because it gives Oscan a very unique feel.

Consonant Stems

In Latin, the consonant stems have almost completely merged with the i-stems, and share the same relation as the o- and io-stems. In Oscan, although partly fused, they are kept much more distinct, and in a future descendent they probably wouldn't fuse. Here are the general endings with notes but I won't give an example just yet as the consonant stems misbehave quite a lot.

SG
Nom ... -s ... -0 ... *-s
Voc ... -s ... -0 ... *-s
Gen ... -eís ... -is ... *-es
Dat ... -eí ... ...*-ai
Acc ... -úm ... -em ... *-em
Abl ... -ud ... -e ... ?
Loc ... -eí ... ... ?
PL
Nom/Voc ... -s ... -ēs ... *-es
Gen ... -um ... -um ... *-ōm
Dat/Abl ... -íss ... -ibus ... ?
Acc ... s ... -ēs ... *-ens

Neuter
SG Nom/Acc ... -s ... -0 ... *-s
PL Nom/Acc ... -ú ... -a ... *-ā

There's a lot of green here. The -úm and -ud endings are from the second declension, the Acc Pl seems to be an innovation and the rest borrow from the i-stems. I'll edit in the rest of this as I'm going on a road trip now, but basically we have 'mute' stems (voiceless stops), nasal stems, s-stems and liquid stems, all behaving slightly differently.

Stop stems

SG
Nom ... meddíss ... liímíz
Voc ... meddíss ... liímíz
Gen ... medíkeís ... liímíteís
Dat ... medíkeí ... liímíteí
Acc ... medíkúm ... liímítúm
Abl ... medíkud ... liímítud
Loc ... medíkeí ... liímíteí
PL
Nom/Voc ... meddíss ... liímíz
Gen ... medikum ... liímítum
Dat/Abl ... medíkíss ... liímítíss
Acc ... medíss ... liímíz

Definitely not sure about the Acc Pl. Also not sure why the double d seems to disappear, but this might be dialectal variation.


r-stems

These correspond solely to either nouns in *-tōr and some nouns of relation.

SG
Nom ... keenzstur ... patir ... maater
Voc ... keenzstur ... patir ... maater
Gen ... keenzstureís ... patereís ... maatreís
Dat ... keenzstureí ... patereí ... maatreí
Acc ... keenzsturúm ... paterúm ... maatrúm
Abl ... keenzsturud ... paterud ... maatrud
Loc ... keenzstureí ... patereí ... maatreí
PL
Nom/Voc ... keenzstur ... patir ... maater
Gen ... keenzsturum ... paterum ... maatrum
Dat/Abl ... keenzsturíss ... pateríss ... maatríss
Acc ... keenzsturs ... paters ... maaters

Not sure at all about the accusative. Also unsure about the epenthetic vowel in patir, which should echo the following vowel but I think would probably level anologically.


n-stems

These correspond mostly to nouns in *-ijōn- and occasionally *-ōn-.

SG
Nom ... legiuf ... humuf
Voc ... legiuf ... humuf
Gen ... legineís ... humuneís
Dat ... legineí ... humuneí
Acc ... leginúm ... humunúm
Abl ... leginud ... humunud
Loc ... legineí ... humuneí
PL
Nom/Voc ... legins ... humuns
Gen ... leginum ... humunum
Dat/Abl ... leginíss ... humuníss
Acc ... legins ... humuns

These correspond directly to nouns like legiō, legiōnis and hōmō, hōminis. I think the first u of humuns should be long as well ie huumuf, huumuneís.
Edit: it's also possible that the nom/acc pl of the first is legiuf and the nom/voc sg of the second is humuns.


s-stems

There isn't really enough evidence although I'll try and put an example together sometime.
Last edited by kanejam on Thu Jul 10, 2014 6:48 pm, edited 1 time in total.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

Fourth Declension

The only certain form in Oscan is kastrúvs, a genitive singular. There are also maním and kastrid, which have i-stem forms. I can't work out whether this might be due the fourth declension collapsing into the third or if only some forms are taken from the third or even if these are just irregular. It probably be possible to reconstruct some forms from the Umbrian and Latin forms but I won't attempt that for now.


Fifth Declension

The fifth declension has fewer forms even than the fourth declension so the same problems exist. I will also leave this as an exercise for the reader.


Adjectives

Adjectives agree with the noun they are modifying in both number, case and gender. As in Latin, there are two classes of adjectives: the first/second declension adjectives, which take the case endings directly from the first and second declension nouns and third declension adjectives, which do the same with the third declension. For example, túvtíks is a first/second declension adjective, and when modifying egmú, a feminine noun, would take the form túvtíkú.


Some Vocab

So far it's all been pretty grammatical, so here are a few words.
túvtú, túvtas - f. people
egmú, egmas - f. thing
húrz, húrteís - m. temple (garden?)
embratur, embratureís - m. emperor
tangiuf, tangineís - m. opinion, feeling
huumuf, huumuneís - m. man, human
memnim, memnieís - n. monument
patir, patereís - m. father
maater, maatreís - f. mother
fraater, fraatreís - m. brother
svesur, svesureís - f. sister
Last edited by kanejam on Thu Jul 10, 2014 6:57 pm, edited 2 times in total.
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

User avatar
Drydic
Smeric
Smeric
Posts: 1652
Joined: Tue Oct 08, 2002 12:23 pm
Location: I am a prisoner in my own mind.
Contact:

Re: Oscan Reconstruction

Post by Drydic »

kanejam wrote:Most of what I have is from reading through Buck's Grammar of Oscan and Umbian. Umbrian is also an interesting language with some neat innovations, but for now we'll focus on Oscan. Buck wrote that 'Oscan is the Gothic of the Italic dialects. In the conservatism and transparency of its vowel-system, it is rivalled only by Greek of all the Indo-European languages. Diphthongs are preserved intact in all positions.'
Just to clarify things, Buck's masterpiece was copyrighted in 1904. The statement quoted is thus pretty dated. This is not at all to say it's a bad book, far from it; to my knowledge, it is still the bible of Osco-Umbrian, and an extremely solidly-written work.
Image Image
Common Zein Scratchpad & other Stuffs! OMG AN ACTUAL CONPOST WTFBBQ

Formerly known as Drydic.

User avatar
Sleinad Flar
Lebom
Lebom
Posts: 124
Joined: Tue Feb 27, 2007 11:18 pm
Location: Coriovallum, Germania Inferior

Re: Oscan Reconstruction

Post by Sleinad Flar »

Partly a bump because I'm eagerly awaiting the conjugations, partly a response on the question of <h>:
I can't find rehtas again, so maybe it was from Umbrian. Here are a list of words that I could find with non-initial h: ehtrad, saahtúm, Úhtavius, feíhuss, kahad, Verehasíúí and also the prefix eh. The h's in the first three are from *kt; the double aa in saahtúm is from 'secondary lengthening' although it doesn't say what that is. Verehasíúí is particularly important because it comes from *werhasjo-, meaning the anaptyxis that is characteristic of Oscan occurs even between *r and *h. The fact that it was transcribed with h in the Latin script doesn't lend much weight as I'd imagine that /x/ would also be transcribed h. The only other spelling that might be suitable would be ch. I'm not sure if there are enough Greek inscriptions with /h/ in them but I'll have a look.

Diachronically, Buck talks of *gh becoming h, with no mention of what happens intervocalically. When describing the phonology, he just transcribes h as h, and again doesn't mention how it might be pronounced non-initially.
Saahtúm apparently is related to sanctum, so the secondary lengthening is from the loss of /n/. Indeed in all those cases, <h> continues older /k/ before /t/. However, there's also piíhiúi , corr. to La. pius < proto-It *p(w)ī-yo-. Apparently <h> separates [i:] from a following or [j] here.
The Wörterbuch mentions no clear etymology for kahad and Verehasiúí (both of uncertain meaning), but feíhúss is from *dhéyg'h-o-, so here at least we have a case of intervocalic -gh- > -h-, whatever the phonetic value.
In view of this evidence, I'm willing to admit <h> represents [x] at least before /t/. I'm still unsure about the other positions.
"Was ist ist, was nicht ist ist möglich"
http://sleinadflar.deviantart.com

User avatar
kanejam
Avisaru
Avisaru
Posts: 257
Joined: Sat Mar 30, 2013 1:16 pm
Location: New Zealand

Re: Oscan Reconstruction

Post by kanejam »

Thanks Sleinad! I've been very distracted but I promise to do verbs next! They'll be a bit trickyAlso thanks for the response on intervocalic h. I found quite a few examples of praenomina and gentiles with an h marking hiatus (or at least what appears to be hiatus) so who knows. I'll have a proper look when I have a minute but maybe h denotes /x/ everywhere except intervocalically, where a previous [*G] (which might be from Proto-Italic to follow the intevocalic voicing of /*f/ and /*s/) has been lost maybe?
If you cannot change your mind, are you sure you have one?

Here's a thread on Oscan.

Post Reply