I am very happy to go into this issue at depth since this has been a main research topic for me since my doctoral dissertation. Right now I start with a short answer: It is not typologically likely that there were contrast between two trills and the phonological activity (as trigger and target) of the other rhotic (written by the yR-rune and descending from PGmc /z/), quite clearly point to an alveolar trill with lowered/non-backed dorsum. In terms of contrast between the two we can be sure of two features: the rhotic written by the reid-rune and descending from PGmc /r/ was backed, since it backed the following vowel AND its was more sonorous as it protected the preceding vowel from syncope. Here is something older I wrote about this. I would adjust the argument somewhat if I wrote it today, but I still concur with the essentials: www.researchgate.net/publication/357807100_The_phonological_properties_of_the_rhotacising_zR-phoneme_in_Ancient_Nordic/references
I already answered this one but now my answer is invisible to myself. Can you see it? In short it is typologically highly unlikely that there were two vibrants, and the yR-sound descending from PGmc */z/ must for a number of reasons have been an alveolar rhotic with at least vibrant allophones. Czech and Spanish have two rhotics, but neither AFAIK contrast them on the basis of "backness", as the Early Old Nordic data requires. From the phonological activity of the two rhotics we may infer what the contrast could have been: the reid-rune rhotic was more "back" (owing to its effect on preceding vowels) and more sonorous (owing to the protection against syncope that it afforded to the preceding vowel, just like /n/, /l/ and /m/ did). From assimilation patterns of the yR-rune rhotic we may infer that it wasn't highly marked (as a palatalised rhotic would be) and it caused fronting (dentalisation) of at least /n/, if not also /l/; so it may have been slightly pre-alveolar. From loanwords we know that it was perceived as a rhotic, not a sibilant.
@JohanSchalin the problem then is that you are requiring a rhotic trill to lenite to an approximant and then re-develop into a trill (I specifically asked for [words like] frą́). On the other hand one can clearly see how sonorous r is when you look at syllable boundaries and lengthening in later scandinavian languages. I am unconvinced but I do think your theory has merit.
@@clanDeCo I do not think that lenition and re-development into a trill is a serious problem, certainly the counter argument is not compelling. First, the developments would have been at least half a millennium apart and would have happened in very different languages. Second, the latter development was not a spontaneous change but a phoneme merger, and needles to say the language users had no memory of what had happened centuries before. Third, the merger need not have resulted in a trill: note that merger would have resulted from rearrangement into complementary distribution and many allophones could have continued as they were, and further many Nordic dialects do not have trills. I fact there is a whole spectrum of more or less backed rhotics.
@@JohanSchalin The dorsal rhotics are fairly late though? I'm pretty sure we have sources that show that both danish and scanian dialects used to have coronal trills, or am I misremembering?
www.researchgate.net/publication/387660111_Christmas_Gospel_Luke_2_1-16_in_reconstructed_Early_Old_Nordic_ca_650-850 Comments and explanations for the text by Schalin.
Yes. This is a reconstructed *_common_* Scandinavian from before that split. Gutnish may have split off before (~650). The colonization of Iceland began in the 870s. Coincidentally, or not, that's when the Scandinavian is held to have split up, but all such processes ought to have been gradual.
The problems in reconstructing this language stage reveal that even at this time depth there were differences between Early Old Nordic dialects. One such problem is the words for "high" and "highest", where descendants of Old East Nordic point to the forms *hɞug-/*hɶyg-, Old Norse point to *hāh-/*hǣh- and Old Gutnish to *hauh-/*hɶyg-. A similar problem concerns the preterite subjunctives: different umlauts seem to apply and umlaut happened in the late 6th century already. On a whole, however, this language is very close to a "common ancestor" of Old East Nordic and Old Norse, and even after this a huge amount of innovations spread to the whole North Germanic language area. so the genealogical tree model is in fact too imprecise to reflect reality.
North Germanic undergoes a kind of "bottleneck" event in the 500s, but different dialectal traits can be clearly dated to the 6-700s onwards. Some traits are hard to date and can also predate syncope, we just don't know for sure.
bsky.app/profile/schaljoh.bsky.social
Tusind tak Johan! Jeg så frem til dine opslag på BlueSky hver dag, og det gjorde min ellers kedelige jul meget mindeværdig!
I'd like to hear his reasoning for reconstructing r as an approximant in frą́
I am very happy to go into this issue at depth since this has been a main research topic for me since my doctoral dissertation. Right now I start with a short answer: It is not typologically likely that there were contrast between two trills and the phonological activity (as trigger and target) of the other rhotic (written by the yR-rune and descending from PGmc /z/), quite clearly point to an alveolar trill with lowered/non-backed dorsum. In terms of contrast between the two we can be sure of two features: the rhotic written by the reid-rune and descending from PGmc /r/ was backed, since it backed the following vowel AND its was more sonorous as it protected the preceding vowel from syncope. Here is something older I wrote about this. I would adjust the argument somewhat if I wrote it today, but I still concur with the essentials: www.researchgate.net/publication/357807100_The_phonological_properties_of_the_rhotacising_zR-phoneme_in_Ancient_Nordic/references
I already answered this one but now my answer is invisible to myself. Can you see it? In short it is typologically highly unlikely that there were two vibrants, and the yR-sound descending from PGmc */z/ must for a number of reasons have been an alveolar rhotic with at least vibrant allophones. Czech and Spanish have two rhotics, but neither AFAIK contrast them on the basis of "backness", as the Early Old Nordic data requires.
From the phonological activity of the two rhotics we may infer what the contrast could have been: the reid-rune rhotic was more "back" (owing to its effect on preceding vowels) and more sonorous (owing to the protection against syncope that it afforded to the preceding vowel, just like /n/, /l/ and /m/ did). From assimilation patterns of the yR-rune rhotic we may infer that it wasn't highly marked (as a palatalised rhotic would be) and it caused fronting (dentalisation) of at least /n/, if not also /l/; so it may have been slightly pre-alveolar. From loanwords we know that it was perceived as a rhotic, not a sibilant.
@JohanSchalin the problem then is that you are requiring a rhotic trill to lenite to an approximant and then re-develop into a trill (I specifically asked for [words like] frą́). On the other hand one can clearly see how sonorous r is when you look at syllable boundaries and lengthening in later scandinavian languages. I am unconvinced but I do think your theory has merit.
@@clanDeCo I do not think that lenition and re-development into a trill is a serious problem, certainly the counter argument is not compelling. First, the developments would have been at least half a millennium apart and would have happened in very different languages. Second, the latter development was not a spontaneous change but a phoneme merger, and needles to say the language users had no memory of what had happened centuries before. Third, the merger need not have resulted in a trill: note that merger would have resulted from rearrangement into complementary distribution and many allophones could have continued as they were, and further many Nordic dialects do not have trills. I fact there is a whole spectrum of more or less backed rhotics.
@@JohanSchalin The dorsal rhotics are fairly late though? I'm pretty sure we have sources that show that both danish and scanian dialects used to have coronal trills, or am I misremembering?
Is this an older form of old norse? What stage is this and when was it spoken?
Finns det en svensk översättning?
Lukasevangeliet kap 2
www.researchgate.net/publication/387660111_Christmas_Gospel_Luke_2_1-16_in_reconstructed_Early_Old_Nordic_ca_650-850 Comments and explanations for the text by Schalin.
Will this split to Old East Norse and Old West Norse? Or was there a "common" Old Norse between this and the Viking Age dialects?
Yes.
This is a reconstructed *_common_* Scandinavian from before that split.
Gutnish may have split off before (~650).
The colonization of Iceland began in the 870s.
Coincidentally, or not, that's when the Scandinavian is held to have split up, but all such processes ought to have been gradual.
The problems in reconstructing this language stage reveal that even at this time depth there were differences between Early Old Nordic dialects. One such problem is the words for "high" and "highest", where descendants of Old East Nordic point to the forms *hɞug-/*hɶyg-, Old Norse point to *hāh-/*hǣh- and Old Gutnish to *hauh-/*hɶyg-. A similar problem concerns the preterite subjunctives: different umlauts seem to apply and umlaut happened in the late 6th century already. On a whole, however, this language is very close to a "common ancestor" of Old East Nordic and Old Norse, and even after this a huge amount of innovations spread to the whole North Germanic language area. so the genealogical tree model is in fact too imprecise to reflect reality.
North Germanic undergoes a kind of "bottleneck" event in the 500s, but different dialectal traits can be clearly dated to the 6-700s onwards. Some traits are hard to date and can also predate syncope, we just don't know for sure.
Gives Old English vibes. Quoth I.