Early old norse.

A forum for discussing linguistics or just languages in general.
Post Reply
User avatar
HinGambleGoth
sinic
sinic
Posts: 432
Joined: 01 Jul 2014 05:29
Location: gøtalandum

Early old norse.

Post by HinGambleGoth »

Since I have been discussing much about old norse with ephraim and Georg standish in the scandinavian forum, I figured an english thread would be good.

The Idea here is roughly to re-construct north germanic as it was spoken at the same time as the "old" west germanic langauges, that is roughly 800-1100, what we call "old" norse is actually mostly contemporary with the "middle" west germanic langauges, old icelandic and old swedish are mostly attested from the 13th century. This is because of the roman-catholic literary culture hadnt reached scandinavia yet, the north germanic speakers were still pagan and had a shoddy writing system using the younger futhark, so getting good contemporary linguistic sources isnt that easy.

The longest runic inscription from this period is the Rök runestone in Östergötland sweden, the language still has some features that were lost i in the manuscript era, such as phonemic nasal vowels, kept short -i and -u, and *z and *r as separate phonemes, this is fairly close to what the danes in the danelaw spoke.

There are some questions though, for instance, was germanic /x/ unassimilated in some norse dialects Well into the Viking age? Some loans into Anglo-saxon seem to suggest this, slaughter, though and tight are listed as :non: loans.
[:D] :se-og: :fi-al2: :swe:
[:)] :nor: :usa: :uk:
:wat: :dan: :se-sk2: :eng:
[B)] Image Image :deu:
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

First something about orthography and historical phonology. I’m going to use a writing system that differs quite a bit from the Old Icelandic orthography. It is perhaps closer to some of the ways that we write Old East Norse, and to some common ways of writing Proto-Germanic. The idea is that the same orthography can be used for all stages of the language from Proto-Germanic, through Ancient Norse (Proto-Norse), throughout the Old Norse Period and perhaps even beyond that. This also means that some distinctions that may have been allophonic at a certain stage are still indicated if they have diachronical relevance. The idea is never to transliterate or imitate a native writing system, it is to somewhat indicate pronunciation.

I don’t really have the ambition of making this a standardized writing system but feel free to use it if you feel like it. If you use another writing system, please specify what you mean by the different symbols.

Most vowel letters have their expected values: <A a E e I i O o U u Y y>. As is common for all historical germanic languages except Old West Norse, a macron and not an acute is used to mark length: <Ā ā Ē ē Ī ī Ō ō Ū ū Ȳ ȳ>.
Short and long /ø/ will be written <Ø ø Ø̄ ø̄>. The i-umlauted short and long a will be written <Æ æ Ǣ ǣ>. The u-umlauted a will be written <Ɔ ɔ Ɔ̄ ɔ̄>. This may confuse readers familiar with Old Icelandic, but the reason for this is that the ogonek has other uses. Also, all vowel letters can be thought of as representing something close to their IPA value. The so called ø₂, evolving from simultaneous i- and w-umlaut of a, will be written <Œ œ Œ̄ œ̄>. Most descriptions of Early Old Norse phonology don’t seem to include this sound as distinct from /ø/, and no description seems to include the long counterpart so it's a topic for discussion whether either of these sounds ever existed.

Overlong vowels of Proto-Germanic (if they existed) will be written with a circumflex <Â â Ê ê Ô ô>. The so called *e₂ will be written <Ě ě>, although I’m not quite sure if we have to reconstruct this phoneme.

The ogonek will be used to mark nasalization. There is a strong tradition of indicating nasalization this way. It’s used in Elfdalian which preserves many nasal vowels, it’s the most common way of writing Proto-Germanic nasal vowels, and <ą> is often used this way when transliterating runes. It also looks better than the alternatives, using a tilde (which doesn’t nicely combine with a macron), an overdot (which almost requires you to use a dotless i) and using a superscript ⁿ. Obviously, it does clash with the familiar uses of <ǫ ę>, especially in Old West Norse, so keep this in mind. For the nasal <Y y>, I will use <Ỿ ỿ> because combining the ogonek with this letter may not look right. There is much more evidence for phonemic long nasal vowels than for short, but I will still indicate both lenght and nasalization separately. The short and long nasal vowels are thus written:
<Ą ą Ę ę Į į Ǫ ǫ Ų ų Ỿ ỿ Ɔ̨ ɔ̨ Æ̨ æ̨ Ø̨ ø̨ Œ̨ œ̨>
<Ą̄ ą̄ Ę̄ ę̄ Į̄ į̄ Ǭ ǭ Ų̄ ų̄ Ỿ̄ ỿ̄ Ɔ̨̄ ɔ̨̄ Ǣ̨ ǣ̨ Ø̨̄ ø̨̄ Œ̨̄ œ̨̄>

I will indicate nasalization even when it was allophonic before nasal consonants. If you want to use the same orthography, this is definetely optional.

Should the need arise, accent can be marked on the syllable receiving primary stress. Monosyllabic accent is indicated with an acute ´ and disyllabic with a grave ` (reconstructing historic accent patterns is obviously not an easy task):
<ɔ̨́ndįn> vs. <ą̀ndįnn>

The familiar three unstressed vowels of Old Norse will mostly be written <A a I i U u>. Their nasalized equivalents will be written <Ą ą Į į Ų ų>. In addition to this, I will (for now at least) write <Ĭ ĭ Ŭ ŭ> for unstressed short i and u that would later be syncopated or apocopated. The unstressed vowel system of Early Old Norse, before the last syncope stage, is a topic for discussion but these must in some way have differed from the regular i and u that was not lost in the same environment. Other stages of the language obviously had different unstressed vowels.

Rising diphthongs (were the stress is on the second element) will be written like a sequence of a semivowel and a full vowel <JA ja JƆ jɔ JŪ jū> etc. Falling diphthongs (where the stress is on the first element) will be written as a sequence of two vowels <IA ia IƆ iɔ EA ea AI ai ÆI æi ƆU ɔu ŒY œy> etc. Nasalization will be indicated on the first element only.

I will write [w] as <W w>, never as <V v>. There is the option of using <V v> for the [v]-realization deriving from older /w/.

As for the consonant, I will write <P p T t K k KW kw> with their expected values. I will write <B b D d G g GW gw> only when it is likely that they had a plosive realization. The voiceless fricatives will be written <F f Þ þ H h HW hw>. The voiced fricatives will be written <Ƀ ƀ Ð ð Ǥ ǥ>. There is no diachronic reason to distinguish between *h realized as [h] and [x], therefore both will use the same symbol. I also don’t see much use in distinguishing between /f/ realised as [f] or [ɸ], although there is the option of using <Ᵽ ᵽ> for [ɸ] or for something that is ambiguous whether it’s an allophone of /f/ or /p/. Similarly, I don’t see any reason to distinguish between the realizations [β] and [v] deriving from historical /f/ and /b/ but there is always the option of using <V v> for the latter.

At most stages of the language, not all of the four series of voiceless plosives, voiced plosives, voiceless fricatives and voiced fricatives are actually distinguished phonemically. But for many reasons, I think it’s sensible to distinguish all if you have a more historical perspective.

I will also write <Ʀ ʀ> for the so called ”palatal r” (as distinguished from <R r>) of North Germanic but <Z z> for the familiar *z of Proto-Germanic. There is actually little diachronic sense in using different symbols (even though I do think there was a shift in realization) so this is mostly keeping with an established tradition.

—————————————

Now, let’s look at some Early Old Norse declension patterns, as I would reconstruct them for now. But please tell me why the reconstruction is wrong!

This may represent a generic East Norse variety spoken in the 8th century. Of course, there was dialectal variation at that time. Notably, Old Gutnish at the time would have had a different realization of the diphthongs. It was already fairly close to later Old Norse varieties, however. Most syncopation and apocopation seems to have already been completed (although there is a distinct possibility that older forms survived in varieties that was never written down). Notably however, original short i and u may have survived after short stems although they were probably in the process of falling.

This is the word *gɔukʀ ’cuckoo’ from PG *gaukaz, and the demonstrative determiner/pronoun *sāsi, which is really just *sā with an added *si (probably from the verb *sēą ’to see’).
*sāsi gɔukʀ m ’this cuckoo’, Old icelandic sjá/þessi gaukr, Modern Swedish denna gök
Nom: *sāsi gɔukʀ — *þæissi gɔukaʀ
Acc: *þąnnsi gɔuk — *þāsi gɔuką
Dat: *þæ̨imsi gɔuki — *þæ̨imsi gɔukųm
Gen: *þessi gɔuks — *þæiʀʀasi gɔuka

Note that all forms of the noun in this declension actually have distinct forms. There is some evidence the accusative plural originally had a nasalized ending (which makes sense as they derive from earlier *–anz). Unstressed nasal vowels were probably lost much earlier than stressed long ones. The nasalization is however preserved in monosyllables in Elfdalian:
http://inss.ku.dk/english/staff/?pure=f ... dalian.pdf

I suspect that feminine ō-stems and adjectives would have had two paradigms based on stem length, like in Old English.

Here’s a short stem adjective and noun.
*swɔlŭ sɔǥŭ f ’a cold saw’, Old icelandic svǫl sǫg
Case: Singular — Plural
Nom: *swɔlŭ sɔǥŭ — *swalaʀ saǥaʀ
Acc: *swala sɔǥŭ — *swalaʀ saǥaʀ
Dat: *swalʀi saǥu — *swalųm saǥųm
Gen: *swalʀaʀ saǥaʀ — *swalʀa saǥa

And long stems:
*tǭm ɔ̨nd f ’an empty spirit’, Old icelandic tóm ǫnd
(remember that old icelandic ǫ/ǫ́ denotes /ɔ(ː)/ while early norse *ɔ̨/ɔ̨̄ denotes /ɔ̃(ː)/)
Case: Singular — Plural
Nom: *tǭm ɔ̨nd — *tǭmaʀ ąndaʀ
Acc: *tǭma ɔ̨nd — *tǭmaʀ ąndaʀ
Dat: *tǭmʀi ąndu — *tǭmųm ąndųm
Gen: *tǭmʀaʀ ąndaʀ — *tǭmʀa ąnda

Now, which forms had u-umlaut I’m very uncertain about. This reconstruction only has u-umlaut on the forms were u was later lost, which is different from Icelandic were u-umlaut was and is still productive.
User avatar
HinGambleGoth
sinic
sinic
Posts: 432
Joined: 01 Jul 2014 05:29
Location: gøtalandum

Re: Early old norse.

Post by HinGambleGoth »

Should the lowering like <is> = <es> "is" be included? Gutnish and Elfdalian lacks this, and you cant really tell from the runic writing either since there was no <e> rune until fairly late.

wesą

jak/ek imĭ
þū ist
þat is
wīʀ iʀųm
þæiʀ iʀų
īʀ iʀuð


That is just present indicative, to be honest, I am wondering about some things, post runic OEN shows an almost complete merger the singular verb forms, with some exceptions, but there are some features that were lost in OWN, such as kept nasal in the 3rd person subjunctive, i guess a combination of Osw and Oic forms would suffice, but neither seem to show 3rd person singular *ð, /ʀ/ *z seems to have been generalized throughout.

Is it justified to reconstruct hann kemĭð ?

how would you render these examples in your orthography?

Þú munt vera Norðmaðr, kenne ec mál þitt.' Þá svaraðe Styrkar,
'Hvat villt þú þá, ef ec em Norðmaðr?


So bygþus þair þar firir. oc enn byggia oc enn hafa þair sumt af waru mali

Sveær egho konong at taka ok sva vrækæ
[:D] :se-og: :fi-al2: :swe:
[:)] :nor: :usa: :uk:
:wat: :dan: :se-sk2: :eng:
[B)] Image Image :deu:
User avatar
DrGeoffStandish
banned
Posts: 581
Joined: 19 Feb 2012 00:53

Re: Early old norse.

Post by DrGeoffStandish »

HinGambleGoth wrote:Since I have been discussing much about old norse with ephraim and Georg standish in the scandinavian forum, I figured an english thread would be good.
Seems like a great initiative! I guess our trio - HinGambleGoth the Energetic Youngster, Ephraim the Scholar and Geoff Standish the Orthography Guy - may make an interesting contribution in the subject.

Some thoughts: In my orthography I prefer to use the conventional acute diacritic in order to denote long vowels. Furthernore, nasal vowels ar marked by adding an apostrophe, this reminds of a lost nasal consonant. (No nasality is marked when the nasal consonant is still present.) The "palatal r" I write with the rune ψ analogous to how the rune þ is used for [θ].

BTW, what's your motivation for your original spelling of the "au-diphthong", Ephraim? It kind of contradicts
      • "The idea is that the same orthography can be used for all stages of the language from Proto-Germanic, through Ancient Norse (Proto-Norse), throughout the Old Norse Period and perhaps even beyond that."
For sure, e.g. Old Icelandic had a completely rounded au diphthong, but a plain au is certainly more archaic.
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

I think there may be two different goals of this thread. One is more descriptive, to discuss all Early Old Norse (as well as Later Old Norse) in general and try to reconstruct what it might have sounded like, based on available evidence (which admittedly, has many limitations). This will obviously take into account that there was a lot of variation at the time, dialectal or otherwise. Some of the differences are in fact recoverable.

A second goal may be more of a conlanging project, to reconstruct a Common Early Old Norse. This would not quite represent a language that was actually spoken (although it probably wouldn't be far from it), it would be a sort of anachronistic combination of the most archaic features from different varieties.

I think both are kind of interesting but what I've written so far relates to the first goal, and the orthography is meant for this purpose. In particular, I wanted to be able to write language changes from PG to ON without changing orthography at different stages. So instead of writing:
*mōdriz > *mø̄ðriʀ > mœðr
Switching from a PG orthography, through an ad hoc late Proto-Norse orthography, to a common Old Icelandic orthography, we should be able to write:
*mōðriz > *mø̄ðriʀ > mø̄ðr

This does not mean that a segments should be represented the same at all stages, it is meant to roughly represent pronunciation. But I do kind of contradict this by using <H h> for both [h] and [x], I guess... In general I wanted to sort of follow established traditions. But I'm not saying it's unflawed and completely principled...
HinGambleGoth wrote:Should the lowering like <is> = <es> "is" be included? Gutnish and Elfdalian lacks this, and you cant really tell from the runic writing either since there was no <e> rune until fairly late.

wesą

jak/ek imĭ
þū ist
þat is
wīʀ iʀųm
þæiʀ iʀų
īʀ iʀuð
I'm not quite sure. Obviously, the late Proto Germanic vowel would have been *i (due to early i-umlaut, *esti > *isti). I don't know what caused the later lowering of *i to *e but I think it's old. Elder Fuþąrk inscriptions would have distinguished the two vowels but I can find only one example of the present copula in Samnordisk runtextdatabas: N KJ60 U dated to 375/400. Here, we find the form ist with *i. This is probably a 3sg form actually, not a 2sg form as you would have expected from Old Norse.

But this is obviously quite a bit older than the Old Norse period so it's inconclusive. I don't know about the Elfdalian forms but the i of Old Gutnish might actually derive from an earlier *e.
Vrieland (Old Gutnish Historical Phonology and the Old Norse Context) p. 20, writes:
"Furthermore, a levelling also occurred in the opposite direction, monophthongizing the 1 and 3 pl (with a raising of *e to i) creating a second, interchangeable paradigm sg *ir,106 *ist, ir, pl *irum, *irin, iru."

It wouldn't sursprise me if the Elfdalian forms are also innovations.

It also wouldn't surprise me if there was a bit of dialectal variation, with both *is and *es. The Younger Fuþąrk would obviously not distinguish them. Perhaps if there are examples of the vowel being written with an a-rune, that would be evidence for a lower vowel. But as far as I can tell (not knowing anything about the Elfdalian developments), the *e-forms are what we find in all later Norse varieties so I'm inclined to reconstruct them for general Early Old Norse.

As can be seen on the old 4th century form ist, the final vowel of Proto-Germanic would have been lost very early on. In general, Ancient Nordic seems to have already lost the final vowel in trisyllabic words, which affected the verb endings. The copula was never trisyllabic, but it was probably often unstressed and there may have been some analogy going on. So I don't think there vould have been any final *ĭ in *ęm. The Proto-Germanic form was actually a heavy stem anyway: *ęmmi > *įmmi.

jak/ek ęm
þū est
þat es
wīʀ/wēʀ eʀųm
īʀ/ēʀ eʀuð/eʀįn
þæiʀ eʀų


I don't know how old the lowering of īʀ to ēʀ is but obviously, ī is the original vowel. I also have no idea when the Old East Norse –įn ending appeared. Due to the nature of runic inscriptions, I don't know if we can really expect 2pl endings there anyway. But the Western –uð ending is original, of course.
HinGambleGoth wrote:That is just present indicative, to be honest, I am wondering about some things, post runic OEN shows an almost complete merger the singular verb forms, with some exceptions, but there are some features that were lost in OWN, such as kept nasal in the 3rd person subjunctive, i guess a combination of Osw and Oic forms would suffice, but neither seem to show 3rd person singular *ð, /ʀ/ *z seems to have been generalized throughout.
Yes, separate 1sg forms were part of Early OEN, I think. But the 3sg *ð seems to have been replaced with *ʀ before the Old Norse Period.
HinGambleGoth wrote:Is it justified to reconstruct hann kemĭð ?
No, I don't think so, not for the Old Norse period. It's not impossible that *ð did survive somewhere but it's not attested as far as I know.

Also, is that a form of *komą? In that case, I suggest hann kømĭʀ for Common Old Norse. The form kømr is found in Old Icelandic, I think kemr may have been a later form actually. It would have been PN *kwęmið > *kǫmið > *kø̨mĭʀ > kømr I think. But forms without the i-umlaut *kǫmĭʀ may have also occured.
HinGambleGoth wrote: how would you render these examples in your orthography?

Þú munt vera Norðmaðr, kenne ec mál þitt.' Þá svaraðe Styrkar,
'Hvat villt þú þá, ef ec em Norðmaðr?


So bygþus þair þar firir. oc enn byggia oc enn hafa þair sumt af waru mali

Sveær egho konong at taka ok sva vrækæ
It depends on if we're trying to represent the language of the period, or if we're translating the language into an older stage. Just representing the language of the period (and being somewhat faithful to the original), I'm thinking:

"þū mųnt wera Norðmaðr, kęnni ek māl þitt." Þą̄ swaraði Styrkar:
"Hwat willt þū þą̄, eƀ ek ęm Norðmaðr?"

(Note the nasalized þą̄. Perhaps this should also include a velarized lateral ł.)

Perhaps this can be made into more archaic Old West Norse:
"þū mųnt wesą Norðmąðʀ, kæ̨nni ek māl þitt." Þą̄ swaraði Styrkārr:
"Hwat wilt þū þą̄, eƀ ek ęm Norðmąðʀ?"

(I'm not quite sure when ʀ shifted to r after dentals (which was before other places), some sources say that it was quite early but there seem to be some ʀ in this position on the Rök stone for example. Also, I'm not sure if *mąðʀ ever had a nasal vowel and if so, when it was lost.)

Sō byǥðus þair þar firir ok enn byggja ok enn haƀa þair sumt aƀ wāru māli.
(I have no idea whether this should include nasal vowels)

Translated into a generic Early Old West Norse variety, perhaps:
Swā byǥðųsk þæiʀ þar fyrir ok enn byggją ok ęnn haƀą þæir sųmt aƀ ǭru māli.
(I think both the forms *wārr or *wą̄rr and *ǭrr may have existed side by side, but I like *ǭrr as it appears to be the regular development from Proto-Germanic. If I remember correctly, it is widely thought that unstessed *ok monophthongized from earlier *auk quite early but that the spelling might have remained on runestones. I don't remember where I read this).

Swēær ēǥo konong at taka ok swā wrækæ.
As for the unstressed vowels, instead of going with the representation found in the manuscripts, they could be simply normalized to <i a u>. I guess it depends on the goal.

This would more archaically be:
Swēaʀ æiǥų kǫnųng at taką ok swā wreką.
(I guess we could write <ŋg> but I don't know if there's any purpose to that).
DrGeoffStandish wrote:Seems like a great initiative! I guess our trio - HinGambleGoth the Energetic Youngster, Ephraim the Scholar and Geoff Standish the Orthography Guy - may make an interesting contribution in the subject.
I like it!
DrGeoffStandish wrote:Some thoughts: In my orthography I prefer to use the conventional acute diacritic in order to denote long vowels. Furthernore, nasal vowels ar marked by adding an apostrophe, this reminds of a lost nasal consonant.
I guess there is a lot of personal preference involved. For now I mostly wanted to be able to write Old Norse and to make it clear what I meant. But for creating more of a Common Early Old Norse, orthography will have to be discussed.

Macron or acute is perhaps a matter of preference, but it's really only for Old West Norse that acutes are conventional (although that is of course were we find the most written material). For almost all other works on older Germanic languages, macrons are conventional (and for other older IE-languages as well). I also like to be able to optionally mark word accent.

Apostrophes for nasal vowels are perhaps an idea. It's certainly much easier to type, but I guess I mostly think it looks wierd. Also, there are some other uses for apostrophes where it marks a lost vowel, 's 'st 'k etc. (for es, est, ek).
DrGeoffStandish wrote:The "palatal r" I write with the rune ψ analogous to how the rune þ is used for [θ].
I quite like the spirit of the idea, I'm not a huge fan of ʀ but it is what's traditional. However, combining Greek and Latin letters looks really wierd in some typefaces. In particular, my favorite typeface for old Germanic is Junicode, which has Greek letters that looks extremly different from the Latin ones.

It also looks very unfamiliar which may or may not be a problem.
DrGeoffStandish wrote:(No nasality is marked when the nasal consonant is still present.)
I definetely agree for Common Early Old Norse. For more descriptive work, it is useful as a friendly reminder to the reader I think. But it's optional.
DrGeoffStandish wrote:BTW, what's your motivation for your original spelling of the "au-diphthong", Ephraim?

[...]

For sure, e.g. Old Icelandic had a completely rounded au diphthong, but a plain au is certainly more archaic.
It is more archaic, but from what I understand, it is thought that the pronunciation with internal umlaut was widespread at the time represented by my example. I'm basing this mostly on Heikkilä p. 110 ff. There is some discussion in the The Nordic Languages, An International Handbook which mentions internal umlaut and the change of *au > *ɔu but I'm not sure if there's an attempt to date the change.

I'm not all that familiar with the reasons for postulating internal umlaut of all diphthongs, but I think it's more that just symmetry, there is some evidence from loanwords into Finnish, for example. There is a lot of logic to the change, though.
*au > *ɔu (internal u-umlaut)
*ai > *æi (internal i-umlaut)
*au > *ɔu > *øy or *œy (internal u-umlaut and external i-umlaut)

It's likely that there were a lot of dialectal variation, though. And in particular, Old Gutnish did not have internal umlaut of *ai, so maybe not of *au either. However, I think the almost universal spelling <au> may actually be partly due to limits of orthography, or perhaps also because many dialects actually returned the diphthong to [au] due to general loss of /ɔ/.

The differences in the realization of diphthongs seem to be one of the variations that are somewhat recoverable and for that reason I especially like to write them differently for different dialects and times. But if the goal is to create a Common Old Norse, I suggest writing <AI ai> and <AU au> as they are more original and more neutral.
Prinsessa
runic
runic
Posts: 2647
Joined: 07 Nov 2011 14:42

Re: Early old norse.

Post by Prinsessa »

Ephraim wrote:This does not mean that a segments should be represented the same at all stages, it is meant to roughly represent pronunciation. But I do kind of contradict this by using <H h> for both [h] and [x], I guess... In general I wanted to sort of follow established traditions. But I'm not saying it's unflawed and completely principled...
If there is no contrast between the two sounds, <h> alone does represent pronunciation perfectly without any sort of contradiction.

Likewise Icelandic today could represent its pronunciation flawlessly with <d> alone as it doesn't contrast with [ð] (orthography might suggest otherwise in some loanwords but those might as well be spelled with <t>), which is how it also happens to be spelled in PG without confusing anybody.

The fewer letters as long as there is no contrast and the difference is allophonic, the better, I'd say. Represent phonemes primarily unless allophony really transforms them somehow, or if a better suited letter is already used for a phoneme elsewhere in the orthography (i.e. write <sagdi> for [ð] and <lamdi> for [d] (i.e. [t] in the case of Icelandic) but <hvæsti> for [t] since /t/ is a phoneme represented by <t> already anyway). Icelandic doesn't represent the exact same plosive-fricative contrast for what I'd refer to as /g/ (i.e. [k]) and [ɣ]), which confuses nobody, so the same could well be done for what I'd refer to as /d/ (i.e. [t ~ ð]).

From what I've seen, Old Norse actually did contrast [ð] and [d] in some positions where Icelandic today does not, or maybe it just had the allophone [ð] where Icelandic today has changed it into [d] (i.e. [t]), like after /l/? Either way, Icelandic was but an example here of how a phonemic representation is often enough.

If the contrast was phonemic in ON, then do write them differently, but it is not in Icelandic, and so there is no reason to have two graphemes there, and the same goes for your <h> [*h ~ *x] or any other similar contrast you might think of, and is in fact exactly the principle used in the conventional romanisation of Proto-Germanic, where <b d g> alone represent both plosives and fricatives according to position, following more or less the same pattern as in modern Spanish, which uses the same orthographical principle, and where <h> too is the sole grapheme for what may have had the same [*h ~ *x] allophony already back then.

The more letters you bring in, the more cluttered and less legible the romanisation will be, so make sure you actually need all those extra letters.

And if you really need <ƀ> then I suggest you use <đ> rather than <ð> for more consistency.
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

The distinction between voiced plosives and voiced fricatives is usually thought to have been marginally phonemic in native words in Old Norse. Specifically it marks word boundaries, and this includes boundaries between the elements in a compound word.

Here’s a ridiculous example:
lāgarni ’billow-yarn-dat.sg’
Lāǥarni ’low-Arni’

The contrast between voiced fricatives and voiceless fricatives also served this purpose (probably). Over time, there was tendency to treat commonly used compound nouns more as single nouns, reducing non-initial vowels and eliminating the plosive/voicing distinction.

Although I should point out that this is based on the most common description of Old Norse, where the contrast between /b d ɡ/ and /ɸ~f θ/ is neutralized into a single series [β~v ð ɣ] between vowels and some other environments, and word-finally, or [b d ɡ] after nasals (or after l for [d]). I think this distribution accounts for the development in all later Nordic languages quite well and it was probably found in Old Icelandic, but we shouldn’t be too quick to assume that all Old Norse varieties had this distribution, especially the older ones. This may be a topic for discussion.
Prinsessa wrote:From what I've seen, Old Norse actually did contrast [ð] and [d] in some positions where Icelandic today does not, or maybe it just had the allophone [ð] where Icelandic today has changed it into [d] (i.e. [t]), like after /l/? Either way, Icelandic was but an example here of how a phonemic representation is often enough.
Yes it seems like the common allophone of /θ/ and /d/ was [d] after /l/ and nasal consonants, but [ð] after /r/. Except in compound words and perhaps in loan words, I don't think the distinction [ð]—[d] was phonemic in any known variety of Old Norse, neither was the distinction [ð]—[θ].

There is something to be said for using separate symbols for the merger of two phonemes, though, rather than arbitrarily picking one. For example, the symbol <ƀ> could be used not so much representing [β~v], but representing the merger of *b and *f.
Prinsessa wrote:The fewer letters as long as there is no contrast and the difference is allophonic, the better, I'd say. Represent phonemes primarily unless allophony really transforms them somehow,

[...]

If the contrast was phonemic in ON, then do write them differently, but it is not in Icelandic, and so there is no reason to have two graphemes there

[...]

The more letters you bring in, the more cluttered and less legible the romanisation will be, so make sure you actually need all those extra letters.
It all depends on the goals. What I described above is not just an orthography for Old Norse, it's meant to also be able to represent Proto-Germanic and Ancient Nordic (Proto-Norse), and perhaps even later stages of language evolution. Specifically, I want to be able to write language evolution (such as sound changes) without switching orthography between the different stages. A switch in letter should always indicate a sound change (the reverse need not be true). Since sound changes operate more on phones than phonemes (although that's a simplification), this means that some non-phonemic distinctions have to be indicated. I could have just used IPA but that would suggest a larger degree of certainty, and would look very unfamiliar.

If you were to design an orthography for only one stage of language evolution, you would of course be more strictly phonemic. But for this to be possible, you have to first reconstruct a phonology and that is actually a topic for discussion. What's a phoneme and what's an allophone changes over time, even without a change in pronunciation.
Prinsessa wrote:exactly the principle used in the conventional romanisation of Proto-Germanic, where <b d g> alone represent both plosives and fricatives according to position, following more or less the same pattern as in modern Spanish, which uses the same orthographical principle, and where <h> too is the sole grapheme for what may have had the same [*h ~ *x] allophony already back then
I don’t think there really is one conventional way of representing PG (there is more for PIE), it seems to be up to the writer. But using plain <b d g> is probably most common (even among writers that differ on other conventions), Ringe uses <b d g gʷ> while Kroonen and Lehmann use <b d g gw>. However, Orel uses <ƀ đ ȝ ȝw> (although yogh ȝ looks almost like ezh ʒ in his typeface, as it does for me on this forum) and he’s certainly not alone. I’ve also seen <ƀ đ ǥ ǥw>. Using <ð> is probably uncommon, though.

The PG realization of the voiced fricatives/plosives is actually subject to some disagreement (and obviously, we have very little evidence to go by so it’s all quite speculative). They may very well have originally been voiced fricatives everywhere and progressively fortified over time (after nasals, after l, when geminated, perhaps initially), and to a different extent in different languages.

Also, Kroonen uses <h hw>, Ringe <h hʷ>, Orel uses <x xw>, Lehmann <χ χw>. The situation is quite messy!
User avatar
HinGambleGoth
sinic
sinic
Posts: 432
Joined: 01 Jul 2014 05:29
Location: gøtalandum

Re: Early old norse.

Post by HinGambleGoth »

Prinsessa wrote:And if you really need <ƀ> then I suggest you use <đ> rather than <ð> for more consistency.
<ð> was an English import that was almost exclusively used in Norwegian/Icelandic and some west götish manuscripts.

Danes wrote <th> and swedes used <þ> for both dental fricatives, realized differently depending on environment, in the younger fuþark <þ> was also used in the same manner.

Actually since the distribution between f/v, þ/ð and g/ɣ was predictable in :non: there is no real need to distinguish them if you want to be economical in terms of letters.

A parallel case is that /ɣ/ was often spelled <gh> in OEN manuscripts, and that is was often written with the hail /h/ rune in runic writing, so it would make just as much sense, from a pan-danish perspective to spell /ɣ/ with <gh> or /VhV/ as it would to use <ð>

An interesting note regarding the letter ð is that it was just a scribal variant of þ in Image, and that its modern use in representations of historical Germanic languages is largely based on its use in modern :isl:
Ephraim wrote: The PG realization of the voiced fricatives/plosives is actually subject to some disagreement (and obviously, we have very little evidence to go by so it’s all quite speculative). They may very well have originally been voiced fricatives everywhere and progressively fortified over time (after nasals, after l, when geminated, perhaps initially), and to a different extent in different languages.
dutch and plattdeutsch(?) g is /ɣ/ i an all positions. High German regularly turned it to /g/ everywhere, and OE and ON only initially, maybe we can assume that this also applied to the other voiced fricatives in the same manner? only earlier.


Speaking of that *ʀ has been a major source of debate, we seem to have a very accurate idea of what :non: sounded like, at least during the literary period, but nobody seems to agree on this sound, I am convinced that is a kind of slack approximant, since it so easily fell of or got assimilated, and also because of the inherent Germanic laziness regarding final sounds.

a part from *ʀ the tonal accent seems to a subject of much debate, one thing i am wondering about is, why is it taken for granted that stød is younger than tonal accent?
Last edited by HinGambleGoth on 18 Feb 2015 20:42, edited 1 time in total.
[:D] :se-og: :fi-al2: :swe:
[:)] :nor: :usa: :uk:
:wat: :dan: :se-sk2: :eng:
[B)] Image Image :deu:
Prinsessa
runic
runic
Posts: 2647
Joined: 07 Nov 2011 14:42

Re: Early old norse.

Post by Prinsessa »

Ephraim wrote:Also, Kroonen uses <h hw>, Ringe <h hʷ>, Orel uses <x xw>, Lehmann <χ χw>. The situation is quite messy!
At least Wiktionary and Ringe agree (and I much prefer this standard myself even tho that's not relevant, but I just wanted to point that out). Who, besides Kroonen, Orel and Lehmann themselves agree on their respective uses?
Last edited by Prinsessa on 19 Feb 2015 10:01, edited 1 time in total.
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

For Proto-Germanic, *z (later *ʀ) is quite unproblematic. It was likely just a voiced variant of /s/, i.e. /z/. The question is when *z became *ʀ, if there ever was such a sound change (or if *z became *r without any intermediate step), and how *ʀ was pronounced. But that's beyond PG.

My favorite hypothesis right now (which I really should stress is very speculative) is that Pre-Proto-Germanic had a system of aspirated voiceless stops /pʰ tʰ kʰ kʰʷ/ that had unaspirated allophones after /s/: [p t k kʷ]. These derive from PIE *p *t *ḱ *k *kʷ. It would also have had a series of voiced stops /(b) t ɡ ɡʷ/ deriving from PIE (*b) *d *ǵ *ǵ *gʷ. Finally, there would have been a series of voiced spirants (non-sibilant fricatives) /β ð ɣ ɣʷ/ deriving from PIE *bʰ *dʰ *ǵʰ *gʰ *gʷʰ. There would have been no unvoiced pairs of the fricatives (exept maybe allophonically), but there would have been an voiceless sibilant /s/. This is not an unusual phonology. Apparently, 10 % of the worlds languages have both voiced and voiceless fricatives but without any pairs contrasting in voice, and in that case spirants are far more likely to be voiced.

The first step of Grimms Law would then have been very similar to the High German Consonant Shift. The aspirated plosives (but not their unaspirated counterparts) would have developed into affricates /p͡ɸ t͡θ k͡x k͡xʷ/ and these would later develop into voiceless spirants /ɸ θ x xʷ/. As a result of there being no voiceless plosives except after /s/, the former voiced plosives would simply devoice to /(p) t k kʷ/. These would merge with the few remaining voiceless stops. The series of voiced spirants didn't change pronunciation, they simply remained as they were.

Verner's Law would then operate on all voiceless fricatives (including /s/), merging them with the old voiced ones unless initial or following stress.

So at this stage, in Early Proto-Germanic, there would not have been any voiced plosives. But I suppose they may have developed already during the Proto-Germanic period, through fortition of voiced fricatives. Especially after nasals and when geminated, this may have happened quite early on (although only after Verner's law).

Again, this is speculative, there isn't really any concensus on quite how the changes happened (or indeed, if Verner's Law was posterior or anterior to Grimm's Law). This is more of a plausible scenario.

I think it's sensible to write *b *d *g *gw for PG even if I suspect (somewhat speculatively) that voiced spirants might have been the primary realization at least for early PG. This is, however, problematic for a diachronic orthography based purely on pronuciation. But it's possible to construct an orthography based more on "diachronic phonemes". We can start to use the symbol *ƀ when *b and *f merge. This is counter-intuitive in one sense, as it doesn't mark a change in pronunciation. But it does mark a sound change. In this case *ƀ is not defined by it's pronunciation but as the product of a merger.

In the same way, the vowel *ɔ is not really defined by it's pronunciation, it's defined as the u-umlaut of *a (however it may have been pronounded).
Prinsessa wrote:
Ephraim wrote:Also, Kroonen uses <h hw>, Ringe <h hʷ>, Orel uses <x xw>, Lehmann <χ χw>. The situation is quite messy!
At least Wiktionary and Ringe agree (and I much prefer this standard myself even tho that's not relevant, but I just wanted to point that out). Who, besided Kroonen, Orel and Lehmann themselves agree on their respective uses?
George Walkden follows Ringe, I think. But due to chance alone, there's always going to be some writers who completely agree on notation!

Wiktionary and Wikipedia seem to mostly base their reconstructions on Ringe, but they don't use quite the same notation. Ringe writes labiovelars with a superscript ʷ whereas Wikipedia uses a normal w (they don't contrast). Also, Ringe uses a double macron for overlong vowels whereas Wikipedia uses a circumflex.

I do agree that the Wikipedia-notation is quite good. It's actually what my orthography is based on (the use of h for [x], circumflexes and ogoneks, and no superscript ʷ).
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

HinGambleGoth wrote:Actually since the distribution between f/v, þ/ð and g/ɣ was predictable in :non: there is no real need to distinguish them if you want to be economical in terms of letters.
It's only predictable if you know where the boundaries in a compound word are. You would normally expect a native speaker to be able to figure it out, but for a language learner it's useful information. If you introduce a hyphen, it becomes predictable: lā-garni vs. Lāg-Arni.
HinGambleGoth wrote:Speaking of that *ʀ has been a major source of debate, we seem to have a very accurate idea of what :non: sounded like, at least during the literary period, but nobody seems to agree on this sound, I am convinced that is a kind of slack approximant, since it so easily fell of or got assimilated, and also because of the inherent Germanic laziness regarding final sounds.
Yes, I also think that a retroflex or alveolar approximant is a likely candidate. I don't think it was fricative as it doesn't seem to participate in voicing assimilation.
HinGambleGoth wrote:one thing i am wondering about is, why is it taken for granted that stød is younger than tonal accent?
It's not really taken for granted or indeed agreed upon I think. I haven't really followed the debate.

It seems that the stød/accent/whatevercamebefore was fixed quite early though, before the definite article became truly suffixed. It's worth noting that the distinction is based on the number of syllables after the last syncope-stage, however.
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

Let's discuss some more archaic forms.

The familiar 1pl pronouns of Old Icelandic are:
nom: vér
acc: oss
dat: oss
poss: vár–r

What we would reconstruct from PG is perhaps something like this:
nom: *wīʀ < *wīz
acc: *ǭs < *uns
dat: *ø̨̄ss < *unsiz
poss: *ǭr–r < *unser–az

I think all those forms are attested. In older Swedish sources we find nominative wir which has a more original vowel. Apparently ǿss is attested in Old Icelandic manuscripts, as is ór–r, which also seems to be the origin of the Elfdalian form. If I'm not mistaken, *ǭs-forms are also attested.

I'm not sure if an actual contrast between the dative and accusative forms are attested, though, it may be that they were conflated early on but that the different forms survived.
User avatar
HinGambleGoth
sinic
sinic
Posts: 432
Joined: 01 Jul 2014 05:29
Location: gøtalandum

Re: Early old norse.

Post by HinGambleGoth »

What is up with the nasal in :elf: wįð ? some kind of contraction?

We have been mentioning old swedish, old icelandic and old gutnish, but what about "Hedeby/Birka norse" and Old dalecarlian? I sometimes come across the theory that :se-da: dialects like :elf: are descended from the viking age dialects spoken in the Folklands. features such as as au = ō h = Ø and ē = ia are shared.

I love this because it shows that :non: wasn't that homogeneous after all, come to think of it, what we refer to as "old Swedish" wasn't even spoken in what today called "Svealand" (the name svealand isn't used until the 15th century, prior to that there where different "folklands" there), i know that Oswe was even called "Gøzku tungu" (geatish tounge) in some manuscript, don't remember exactly where though, and "a suensku" ( in swedish) isn't used until the mid 14th century if not later another interesting fact is that the people in mälardalen arguably still had au and æi since they brought those pronunciations with them when they colonized Norrland and Finland during the high medieval period, neither are found in the written standard based on speech in östergötland at the time.

So actual old swedish, not the old east-geatish spoken by guys like Birghir iærl would be an interesting topic.

Speaking of Old Swedish, what where the values of the ending vowels after long syllables? a = æ, i = e, u = o should this be accounted for in the reconstruction?
Ephraim wrote: The first step of Grimms Law would then have been very similar to the High German Consonant Shift. The aspirated plosives (but not their unaspirated counterparts) would have developed into affricates /p͡ɸ t͡θ k͡x k͡xʷ/ and these would later develop into voiceless spirants /ɸ θ x xʷ/. As a result of there being no voiceless plosives except after /s/, the former voiced plosives would simply devoice to /(p) t k kʷ/. These would merge with the few remaining voiceless stops. The series of voiced spirants didn't change pronunciation, they simply remained as they were.
i was thinking about a re-conlang based on this idea at one point. You know, all those (very) early loans into uralic, made me think, what was the rest of the language like?
[:D] :se-og: :fi-al2: :swe:
[:)] :nor: :usa: :uk:
:wat: :dan: :se-sk2: :eng:
[B)] Image Image :deu:
Prinsessa
runic
runic
Posts: 2647
Joined: 07 Nov 2011 14:42

Re: Early old norse.

Post by Prinsessa »

Ephraim wrote:Wiktionary and Wikipedia seem to mostly base their reconstructions on Ringe, but they don't use quite the same notation. Ringe writes labiovelars with a superscript ʷ whereas Wikipedia uses a normal w (they don't contrast). Also, Ringe uses a double macron for overlong vowels whereas Wikipedia uses a circumflex.
Yeah, but those are very minor and not really a difference in letters themselves. I do enjoy those double macrons and the superscripts tho. Pretty used to both as Ringe and Wiktionary are my primary sources of contact with PG. Maybe I'm just biased. c; But I do seem to see pretty much these conventions (whatever minute differences there may be) pretty much everywhere I look for PG, so it does feel more like the standard to me than other, more clearly different, conventions.
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

HinGambleGoth wrote:What is up with the nasal in :elf: wįð ? some kind of contraction?
There is an article in Swedish by Lars Steensland on the topic of parasitic (unetymological) nasalization in Elfdalian:
http://uu.diva-portal.org/smash/get/div ... FULLTEXT02
p. 116

"Som alternativ till Noreens och Levanders förklaring att wįð ~ wįr skulle ha uppstått genom ”morfologisk analogi” vill jag föra fram hypotesen att pronomenet fått sin nasalitet från verbändelsen -ųm, vilken mycket ofta föregår pronomenet (t.ex. irųm wįð ’är vi’). Det finns exempel på liknande företeelser i flera andra språk.91 Här vill jag påminna om hypotesen att įð ~ įr skulle ha fått sin nasalitet från den föregående personändelsen -in.92"

This was actually my spontaneous hypothesis as well. There is of course a strong tendency in the Nordic languages to borrow the final consonant of verbal endings as the first consonant of pronouns. Compare Swedish ni, Icelandic þ and most analoguously Nynorsk me. Elfdalian didn't borrow the consonant but it's likely that there was some influence. The language seems to have a tendency to nasalize vowels after nasal consonants, not just before.

Some other unetymological nasal vowels are harder to explain.
HinGambleGoth wrote:Speaking of Old Swedish, what where the values of the ending vowels after long syllables? a = æ, i = e, u = o should this be accounted for in the reconstruction?
I don't know what phonetic value they represented, if they were close to the value of the stressed vowel or if they actually represented a more reduced value. But vowel balance and vowel harmony is something that you could choose to represent if you know it was part of a certain dialect, or choose not to as it's predictable or mostly predictable and seems to be subject to a lot of variation. When vowel balance and vowel harmony appeared, I don't know but the fact that syncope and apocope depended on the length of the stressed syllable may indicate that something akin to vowel balance was part of the Nordic languages from an early age. Or at least that non-initial syllables were stressed differently depending on the length of the initial syllable.
User avatar
HinGambleGoth
sinic
sinic
Posts: 432
Joined: 01 Jul 2014 05:29
Location: gøtalandum

Re: Early old norse.

Post by HinGambleGoth »

Ephraim wrote:When vowel balance and vowel harmony appeared, I don't know but the fact that syncope and apocope depended on the length of the stressed syllable may indicate that something akin to vowel balance was part of the Nordic languages from an early age. Or at least that non-initial syllables were stressed differently depending on the length of the initial syllable.
This is an interesting subject, since vowel balance working much the same as in old Swedish, also appears in Ingveonic old frisian, maybe it was one of those "hedeby" features?

Ofris is pretty interesting since it shares alot of features with ON, like nasal assimilations uns = us , a-stem plural as "ar" "dagar", w-breaking *singwan = siunga.
[:D] :se-og: :fi-al2: :swe:
[:)] :nor: :usa: :uk:
:wat: :dan: :se-sk2: :eng:
[B)] Image Image :deu:
Prinsessa
runic
runic
Posts: 2647
Joined: 07 Nov 2011 14:42

Re: Early old norse.

Post by Prinsessa »

HinGambleGoth wrote:
Ephraim wrote:When vowel balance and vowel harmony appeared, I don't know but the fact that syncope and apocope depended on the length of the stressed syllable may indicate that something akin to vowel balance was part of the Nordic languages from an early age. Or at least that non-initial syllables were stressed differently depending on the length of the initial syllable.
This is an interesting subject, since vowel balance working much the same as in old Swedish, also appears in Ingveonic old frisian, maybe it was one of those "hedeby" features?

Ofris is pretty interesting since it shares alot of features with ON, like nasal assimilations uns = us , a-stem plural as "ar" "dagar", w-breaking *singwan = siunga.
Definitely seems to have a lot of NG-looking and less WG-looking <j>'s everywhere to this day. What are the modern plurals like?
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

Apparently, vowel balance is commonly thought to have originated as a contrast in quantity and not in quality. Vowels were lenghtened after short stems. There was of course a general tendency to lengthen older bimoraic CV.CV-words to three moræ, and logically that leaves three options (the first two are more familiar):
CV.CV > CVC.CV
CV.CV > CVV.CV
CV.CV > CV.CVV

This would have shifted to a difference in quality later but there are still a few Norwegian and Swedish dialects where vowel balance involves a difference in quantity (in addition to a difference in quality).
http://edoc.hu-berlin.de/skandinavistik ... smenko.pdf

It seems that evidence for vowel balance appears only in the 14th century (so not exactly Early Old Norse) and it seems to be an exclusively Northeast Scandinavian phenomenon. I don't know if this could juust be a lack of good source material. If the original difference was one of quantity, I suppose it’s less likely to have been indicated i writing.

Vowel balance can of course be related to the tendency in the history of Germanic languages to treat vowels differently after light and heavy syllables, but vowel balance takes it a step further. It’s not only reduction after heavy syllables, it’s also strengthening after light syllables.

If there’s any relationship to Frisian vowel balance, I don’t know.
User avatar
HinGambleGoth
sinic
sinic
Posts: 432
Joined: 01 Jul 2014 05:29
Location: gøtalandum

Re: Early old norse.

Post by HinGambleGoth »

I have wondered about some things, i have sporadicly been reading an old antiquarian edition of the Östgötalag, trying to gradually understand :non: i have been noticing some interesting spellings.

ending vowels are consistently written, compared to other Oswe manuscripts that often show vowel balance or weakening, <i> is often lowered to <e> particulary in dative singular, but it is not consistent, and seemingly not dependant on syllable weight.

/ia/ and /io/ arent fronted, this is otherwise a typical Oswe shift.

there are no /j/ written between velars and front vowels, like in most other Oswe manuscripts.

*R is almost randomly dropped finaly.

I find these features interesting since the law is considered quite young, dated to 1290. Maybe it is an edited older law? what we know for sure is that Birghir iarl is mentioned.
[:D] :se-og: :fi-al2: :swe:
[:)] :nor: :usa: :uk:
:wat: :dan: :se-sk2: :eng:
[B)] Image Image :deu:
Ephraim
sinic
sinic
Posts: 386
Joined: 15 Nov 2013 13:10
Location: Sweden

Re: Early old norse.

Post by Ephraim »

HinGambleGoth wrote:I have wondered about some things, i have sporadicly been reading an old antiquarian edition of the Östgötalag, trying to gradually understand :non: i have been noticing some interesting spellings.

ending vowels are consistently written, compared to other Oswe manuscripts that often show vowel balance or weakening, <i> is often lowered to <e> particulary in dative singular, but it is not consistent, and seemingly not dependant on syllable weight.

/ia/ and /io/ arent fronted, this is otherwise a typical Oswe shift.

there are no /j/ written between velars and front vowels, like in most other Oswe manuscripts.

*R is almost randomly dropped finaly.

I find these features interesting since the law is considered quite young, dated to 1290. Maybe it is an edited older law? what we know for sure is that Birghir iarl is mentioned.
I don't know that 1290 is that young in Old Swedish terms, it's still well within the Early Old Swedish period. Although Östergötland seems to be within the area that had vowel balance, 1290 is actually just before the first traces appear in manuscripts.

The so called progressive i-umlaut of short ia as in stiala first appeared already in the 11th century according to Wessén, but <ia> still appear in manuscripts after that so it's not that surprising if it's still found in the 1290. The other progressive i-umlauts, of , , jo (I assume that they have shifted to rising diphthongs at that point) apparently don't appear until the 14th century.
Post Reply