Talk:Old Norse morphology

ERROR: Concerning "hit" and the (re-)grammaticalization of the definite article
As far as I can see there is a blatant error in this section. The past participle of strong verbs, i.e. the ones without dental suffix, has NOTHING TO DO HERE. There is merely a homonymy in the declination (i.e. they look/sound alike). The grammaticalization (suffixation) of hinn/hit is something completely different; consult any old norse grammar. I tried to change the table to exclude the forms of "kominn", which should really be removed in order not to confuse. Could anyone please do this to assuage my distress? I sadly do not have the know-how to remove these forms myself; I have never really edited tables at Wikipedia. In advance: thanks!

Additionally: I am quite certain that this is wrong as well:
 * "The first form of the definite article was et/enn/en.[12][25] It was originally a distinct word, placed after the noun.[12] Later, it appeared as it/inn/in, and in its free form also as hit/hinn/hin.[V 16]"

First of all, the sources are quite suspect, to say the least. Secondly, I am quite certain it is exactly the other way around, i.e. hinn (pre- and postpositioned) > inn (postpositioned) > -inn (encliticized and later grammaticalized). (to lend some credence to my argument, I should mention I have an MA in Old Norse, even though it is tacky "degree-dropping"). Furthermore, I am positive that the forms with /e/ (e.g. enn) are later forms, with a vowel reduction ([i] > [e]) induced probably by high token frequency. Sadly I do not, at the moment, have a source for these things at hand, but it should not be to difficult to procure as soon as I get home. In the meantime, or if I should come to forget, could anyone please help out? I am hesitant to change or delete without a proper source at hand.

--Alexlykke (talk) 10:34, 5 January 2014 (UTC)


 * As for past participles, your are correct that the etymology is different. However, in substantive uses of the past participle, the part of speech and declension are identical. So it is also correct to say they are the same declensions. This is the difference between a diachronic and synchronic analysis of the morphology. The modern English words hand and arm are of the same declension, even though they have etymologically different declensions. As for enn and hinn, I agree. It's also inconsistent with the earlier paragraph. ᛭ LokiClock (talk) 09:59, 22 August 2014 (UTC)


 * I changed the title so the section is about the declension pattern, not the word hit, and characterized past participles on their own. More needs to be done. ᛭ LokiClock (talk) 10:48, 22 August 2014 (UTC)


 * On the other hand, maybe it would be better if it were just about hit and its derivatives, since what's special about those words is not their declension, which is just like any other adjective, but the properties of the article. The strong verbs section already states, "The past participle of strong verbs follows the pronominal declension of hit,[V 4] though unlike the definite suffix the participle is inherited from Proto-Germanic." All that lacks is a characteristic translation an adjectival past participle of a verb could have. ᛭ LokiClock (talk) 11:58, 22 August 2014 (UTC)


 * Sleeping on it, the section is intended to be about hit, and so it is as you see it, komit only there because the article means to say the suffix is the same. So I'll put it back to what it was and remove the table column you wanted removed. ᛭ LokiClock (talk) 22:17, 22 August 2014 (UTC)


 * On the other hand, the past participle table here:


 * has bisyllabic stem type declension, which is difficult to infer. Could replace gótt as an example of the strong adjectival declension, if it wouldn't make the table too wide. Would it be enough to say in Strong verbs it is a bisyllabic stem which contracts like hamarr?


 * By the way, aside from removing the rows you need gone, the only thing you need to do to remove a column group from a table in the way I've written them is to look at where it says colspan on the title (Possessive Pronouns) of the table, and take a way from that whatever number is on the colspan of the title (mitt (mín)) of the column group you wanted to remove. Removing a column from a column group is just one layer deeper - lower the colspan of the table, lower the colspan of the column group, and remove the column's rows. ᛭ LokiClock (talk) 22:44, 22 August 2014 (UTC)

To-Do List
From the main article. ᛭ LokiClock (talk) 20:58, 21 March 2010 (UTC)

Information required to document these items may likely be mined from texts that have already been referenced in the article. ᛭ LokiClock (talk) 05:21, 2 October 2010 (UTC)


 * 1) Nouns
 * 2) Strong declensions - Feminine and masculine i- and r-declensions. Expertise in OEN declension appreciated.
 * 3) +Negative suffix
 * 4) Adjectives/Pronouns
 * 5) +Demonstratives (sá, þessi, hinn/hitt)
 * 6) Definite article: Note on limited usage and rarity in verse due to awkwardness of meter.
 * 7) +Adverbs
 * 8) Verbs
 * 9) Redone present-preterite & suffix tables using the format of the strong and weak tables
 * 10) +Personal suffix
 * 11) Weak verbs: Derivation of dental suffix from an obsolete cognate of "to do," according to Cleasby-Vigfússon (Remarks on Strong, Irregular Verbs)

Dialectal morphology
A notable complaint about Old Norse, largely I feel due to the sources for morphology, is that it is or was biased towards OWN, particularly OIC. Besides noting which paradigms merged in the dialects, what are some suggestions for improving the bias? I think we should go with neutralized versions of earlier Old Norse morphology and leave the ca. 13th-15th century tables for the individual languages' articles. One example of a neutralization would be using Ʀs in their proper locations, because the Western paradigms are easily inferred from this. Another would be using the diphthongs in the ablaut patterns, rather than the Eastern monophthongized vowels. Obviously, if a paradigm is drastically different, extra tables would be created, rather than distorting the paradigms beyond application to any of the dialects. ᛭ LokiClock (talk) 03:43, 23 March 2010 (UTC)

Strong neuter tables
I've created these in my sandbox, but I'm worried about putting them up because the division of the stems into A-like and I-like is my doing, and the identification of nęs, klæði & ríki with the yo stems and iyo stems. In addition, I don't know the stem classification of kǫtt-like masculines, which hǫgg would, I further assume, be identified with. I do, however, have explicit identification of barn and tré with o (armr) and u (sonr) stems. But if my intuitions can be supported by further research, I'd hate for the tables, and the proposition of this division, to remain completely unmentioned. ᛭ LokiClock (talk) 00:28, 25 July 2010 (UTC)


 * Neuters have been added. I see no other o stems for hǫgg and the like to share ancestry with. However, I cannot post some information on the feminine A declension without first confirming the class of nouns in heiðr, ęrmr, &c. which I suspect to be iyā. ᛭ LokiClock (talk) 11:03, 8 October 2010 (UTC)


 * Worked around by referencing in terms of the Norse stems, but those PIE/PGmc stems need to be added anyway for complete coverage. ᛭ LokiClock (talk) 12:12, 8 October 2010 (UTC)

Old Norse morphology, removed information
(copied from --91.148.159.4 (talk) 14:25, 24 January 2012 (UTC)'s talk page)

Hello. The information on the gender of the past participle was sourced. Although there were weak verbs with an unrelated -inn, which you might have been thinking of, the past participles in -inn are in fact the article, and past participles are declined for gender whether they're weak or strong. In the future, if you question sourced content, please read the sources, and if a claim is not represented in the source tag it with check source or citation needed. ᛭ LokiClock (talk) 12:49, 24 January 2012 (UTC)


 * First, the link to the source is dead and not preserved on the Internet Archive, so I can't read it (and can doubt that it has been interpreted correctly). Second, the past participles in -inn are not "in fact the article". They have nothing to do with the article historically, they have no definite or pronominal meaning as it does (to produce their sort-of definite form, you have to decline them with the special weak endings). They existed in Proto-Germanic long before the suffixed article arose in Old Norse (e.g. *kumanaz for kominn). So no, the verbal and adjectival endings in -inn are in no way definite, and conversely, it is strange to say that the definite nouns "are turned into" adjectives just because the definite article declines with strong adjectival endings: the nouns still function syntactically as nouns and they also preserve their nominal endings before the article. And by the way, I never wrote that weak participles don't decline for gender.--91.148.159.4 (talk) 13:58, 24 January 2012 (UTC)


 * I have now read the sources on the other site, which isn't dead (http://lexicon.ff.cuni.cz/), and as far as I can see, the source does not claim anywhere that the ending of the strong verbs "is" the definite article (nor does it equate definite nouns with adjectives). As for policy practice, I'm not obliged to use those tags rather than remove the information immediately - the burden of proof (WP:BURDEN) is on those who want to keep the information. In this case, I felt that the claim merited immediate rewording - in any case, by saying that the two are phonetically identical, I preserved your basic insight, just without the historically misleading wording that equates the two things.--91.148.159.4 (talk) 14:22, 24 January 2012 (UTC)


 * Yes, he just groups them together (as adjectives in -inn) and says the morphology of the participle conforms to the article, which was the real information. I must have read into that. Now the mistake seems obvious, as the other languages have the same participle. I know you're not obliged to use tags, that's why I said please. When the source is a dead link, that doesn't mean the source ceased to exist, and thus that the statement didn't have support to begin with. If the link is dead, there's a dead link template which will alert editors who are familiar with the source to the missing information and allow them to restore it. ᛭ LokiClock (talk) 23:03, 24 January 2012 (UTC)


 * OK, I'm glad that we agree, all in all. It's true that the author says of the -inn participle that "with the exception of the contraction it conforms to the article". This basically means that it coincides with it in its phonetic form. And of the article he says that it "forms a double declension, with substantive and adjective forms in the same word; or rather it gives to the substantive the form of an adjective". Which is, again, a way to say that the endings of the article coincide phonetically with those of strong adjectives. These remarks are appropriate and useful for practical pedagogical purposes, to help learners memorize the morphology, and it is fine for them to be reflected in the article in some way. What I objected to is the direct equation of the things, ignoring their semantic, syntactical and historical aspects.
 * About dead links - sure, they still count as sourcing, at least when it's clear what publication they were supposed to point to, and of course, it's also perfectly OK to have just a citation with no link to anything at all. But I was quite confident that this wasn't really about the facts of Old Norse described in the source, but about their wording; and I was also guessing that the wording had shifted a bit, unconsciously, in comparison with the source, as indeed turns out to be the case. And of course, even if the wording hadn't shifted, wordings found in a source from 1874 can be in need of improvement even if the factual claims are reliable. Best, --91.148.159.4 (talk) 15:23, 26 January 2012 (UTC)

Comparative coinciding with the agent noun
"The comparative form as in Latin, other Germanic languages etc. is also corresponding to the agent noun-ending, which in Old Norse has a weak declension with roots in -an-stem inflection as well."

This observation feels a bit too amateurish and it would be best to skip it, IMO. There is no historical connection between the comparative and the agent-noun. The agent nouns are not weak in the other languages, so the similarity is pretty much restricted to the nominative. Even in Old Norse, the paradigms do not coincide, because the adjective weak declension does not coincide with the weak declension of nouns in much of the paradigm. The coincidence is not inherited from Proto-Indo-European, so the Latin comparison is irrelevant, and the Latin comparative ending -ior does not coincide exactly with the agent suffix -or in any case. I'm not deleting it mostly because I am pathologically reluctant to risk being involved in an emotional argument and edit conflict with the author of the sentence. --Anonymous44 (talk) 19:28, 30 May 2014 (UTC)


 * Here is the Diff. It's uncited, so it's best to remove it. The original wording is slightly different, in that the nominative is referred to, not the agent noun-ending, which is less accurate in that both genitives and datives can be agentive cases for certain verbs. I think I wrote the original statement that the cases were modeled after the weak nouns, which is wrong by your research. It should then say they follow the weak nouns, but that this parity isn't etymological. That way the synchronic analysis of the grammar can be preserved - which is important because it could be a merger that happened because of pattern recognition and analogy. ᛭ LokiClock (talk) 03:47, 31 May 2014 (UTC)