Wikipedia talk:Naming conventions (Old Norse/Old Icelandic/Old English)/Archive 1

Proposal
In articles dealing with Old Norse mythology the spelling of names should be that of standardized 13th century Old Icelandic as found in, for example, the Íslenzk fornrit edition of Snorri Sturluson's Heimskringla.

In accordance with the English Wikipedia convention to use forms familiar to English speakers the following exceptions are made:

1. The English familiar form Thor should be used instead of the Old Icelandic form Þórr.

2. The English familiar form Odin should be used instead of the Old Icelandic form Óðinn.

3. In other cases when someone proposes that a familiar English form exists, this can be discussed on an individual basis on the relevant talkpage.

Due to current technical limitations the letter oe-ligature (&oelig;) cannot properly be used in article titles. For purposes of article titles it should be replaced with 'oe' while &oelig; can used throughout the article. Since o-ogonëk will not display correctly on many systems it should be replaced throughout with 'ö'. This is unambiguous since 'ö' does not otherwise occur in standardized Old Norse spelling, and it is a common replacement for o-ogonëk.

Reasoning
As described in Old Norse orthography there is currently no generally accepted standard for the use of Old Norse names within an English text, inside or outside Wikipedia. One source may use the anglizised form Balder while the next employs the modern Icelandic form Baldur and the third uses the standardized Old Icelandic form Baldr. The mainland Scandinavian languages have their own familiar forms of the names and those sometimes also make their way into English prose. Current examples on Wikipedia include the Swedish form 'Väderfölni (seen on Norse mythology) and the Norwegian form Vêrfolne (seen on Vêrfolne) for the being known in Old Norse as Veðrfölnir.

It is the hope of the authors that this standard establishes a consistent and accurate convention to use within Wikipedia.

Examples

 * Loki
 * Baldr
 * Freyja
 * Freyr
 * Heimdallr
 * Angrboða
 * Svaðilfari
 * Höðr
 * Þrymr
 * Veðrfölnir
 * Hoenir as article title, H&oelig;nir within the article

Activation
This standard was proposed by Wiglaf and Haukurth on May 21st 2005. It will become active one month from that date if no objections have surfaced by then.

=Old discussion below=

The articles on Norse mythology show a variety of forms: Old Norse, Modern Scandinavian and modern Icelandic.

The main page says: Generally, article naming should give priority to what the majority of English speakers would most easily recognize with a reasonable minimum of ambiguity, while at the same time making linking to those articles easy and second nature.

Concerning names such as Loki, Odin and Thor, there are standard forms. You will however find other forms that may seem confusing for some less knowledgeable readers, such as Þórr. such forms should be standardised (within articles) to make the reading easier. A limit to this standardisation should be when there are non-Scandinavian forms, such as Thunor or Woden.

Should there be a rule, such as


 * 1) If there is a standard form in English, then that form is used.
 * 2) If there is no such standard form a modern Scandinavian form should be used (I don't mean this as an offense to Icelanders but Wikipedia is a work of reference and there are about 80 times more Scandinavians than Icelanders.)
 * 3) If there is no modern Scandinavian form, the Old Norse form is used.

Any opinions? Wiglaf


 * Here's one opinion. I agree with the above. Personally, I wouldn't dream of imposing modern Icelandic on the world. :-) But I have a question about:


 * If there is no such standard form a modern Scandinavian form should be used.


 * Which modern Scandinavian? AFAIK, there's no standard form for most of the names, and there are at least four Scandinavian languages (here meant to be synonymous with continental North-Germanic). Is there some work in the field which can be supposed to be known to most educated Scandinavians? If so, I'd suggest using that.


 * In any case, I support Wiglaf on this with the following addition:


 * If names from mythology are given in an English or Scandinavian form, it should be permitted as a matter of course, without anyone taking offence, to add the standardized Old Norse names in parentheses where appropriate. See Baldur and Valkyrie for examples of what I mean. Furthermore, that the Old Norse names, where supplied, should conform to the standard as applied in the standard Icelandic scholarly editions (samræmd stafsetning forn). I'm not suggesting that out of nationalism, but because, AFAIK, that is the only standard there is. Editors outside Iceland largely follow their personal conventions. Cheers Io 16:33, 21 Jul 2004 (UTC)


 * PS: I'm wondering, since Wikipedia is a work of reference, whether it might be a good idea to create a page called Norse_Mythology_(names), where the English or Scandinavian names are listed with their Old Norse equivalents. Readers want to read about who Oden/Odin/Óðinn was of course, but they might also want to know what he was actually called. I'd be willing to this, if it is believed to be woth the while. Any comments? Io 16:47, 21 Jul 2004 (UTC)


 * I'll do it. It will take a while, but I think we have a solution to make everyone happy. And the best of luck with your thesis. Io 20:33, 21 Jul 2004 (UTC)

Certainly my understanding of the correct English spelling of Baldur was, er, Baldur. But let us not trivialise or demean the deities by transliterating them slavishly; let us have our Jords, Manis et al. It retains the charm, allure and mystique of a charming and beautiful mythos. Sjc 19:49, 21 Jul 2004 (UTC)

More Information
No, particular site that I know of that uses Cassell especially. Everyone's use is probably slightly different.

Joseph Campbell used no diacritics and ae for æ, at least in his popular writing. He similarly generally avoided fancy diacritics and spellings for Hindu or Semitic names.

In Poul Anderson's War of the Gods he used no diacrtics and æ as ae. But in Poul Anderson's more recent novel Mother of Kings though he continued to use ae rather than æ and no acute marks, o-ogonek or o-dieresis was represented by ö. As I mentioned previously, the practise of dropping the acute accent but retaining ö has been a common one. This seems to be what the web Encyclopædia Britannica mostly intended to do, as it has forms such as Jörd, Njörd, Jörmungand, Höd, Mjölnir, Ragnarök, and Völuspá. The use of á here is unusual for Britannica. I don't find any other acutes in Norse forms except where it is specifically giving the genuine Old Norse forms as opposed to its Anglicized versions. Another recent example that retainins only ö of vowels with diacritics is Barbara Leonie Picard's Tales of the Norse Gods.

An old example of this method of Anglicization is found in H. A. Guerbir's Myths of the Norsemen from the 1920s. (Guerbir also generally uses i instead of j, something not much found any more though not entirely dead.).

The humorous Norse Mythology ... according to Uncle Einar drops all diacritics but keeps æ, following the Penguin norm.

Microsoft Encarta drops all diacritics. But it lists Freyr as Frei and Freya as his wife!!! See Encarta: Frei. People pay for this?

On the other side, John Lindow's Norse Mythology: A Guide to the Gods, Heroes, Rituals and Beliefs (which contains a short dictionary/encyclopedia of Norse gods and heroes) uses full diacritics (only replacing o-ogonek by ö), using either the same system as the Cassels's Dictionary of Norse Myth & Legend'' or something very close to it. I presume the same method appears in Handbook of Norse Mythology by the same author. Jean Young's translation: Prose Edda of Snorri Sturluson: Tales from Norse Mythology from 1954 does the same, except that even ð is retained, so Óðnn is just 'Óðin.

For the ultimate extreme, The Oxford History of the Vikings contains only Old Norse and other original forms, no Anglicizations whatsoever.

I may be overdoing it, but I would like a clear record as to the justification for whatever system is decided on.

I personally agree with Wiglaf on generally retaining diacritics, even though they are more often dropped in translations or appearances in modern fictional works, because an encyclopedia is in a somewhat different category, the diacritics are not always dropped in English translations, and scholarly works of a popular sort often retain them. However one would still often want to list the actual supposed standard Old Norse form to provide case ending information.

To play devil's advocate, since Anglicized forms using only strict US-ASCII spellings are so very common in English use (including many scholarly translations) a case could be made that generally only two forms should be listed, the extremely Anglicized form (as the norm) and the full standard Old Norse form with diacritics, this form to be listed in the header paragaph of names that have their own entry and on the first mention in an article for other names. From these two forms all other forms can be extrapolated, though where in common use they might be listed separately.

For example, if fully Anglicized forms are standard, then we have articles entitled Aesir and Njord and Thor which mention in the header paragraph that the genuine Old Norse forms are Æsir and Nj&#491;rðr and Þórr. If Anglicized forms with diacritics and with æ retained are to be standard, then we have articles entitled Æsir and Njörd and Thór but with the same mention of genuine Old Norse forms. (For a few names like Thor, where a particular form is very common, we would however make an exception as is done by Brodeur and by Cassell's.) The reader is to understand that diacritics are often removed. But the numerous alternate forms can be and probably should be listed by template, for example: Njörd (Old Norse Nj&#491;rðr), the Norse god of .... etc. etc. Alternate renderings in English: Nr(r) Or is this listing superior?: "Other renderings in English: Njord, Niord, Niörd, Njordr, Njördr, Njörð, Njördhr, Njordhr, Njorth, etc." Both rendering methods are messy, yet all forms given here in the second method are found in English for this one single-syllable name. Perhaps use the template format only when there are more than four possibilities. Many have far less, for example Balder has only three: Baldr, Balder, and Baldur. But that could be represented by the template Bald(e/u)r.

I set forth here what seem to be the common ways in which English forms should differ from standard Old Norse forms in extreme Anglicization:
 * 1) The letters þ, ð and &#491; should be represented respectively by th, d, and ö (not considering in this point whether ö should be retained as ö or changed to o which is the subject of point 4). Y/N
 * 2) Nominative case endings should normally be dropped (setting aside for the moment any exact rules about when they might be occasionally retained). Y/N
 * 3) The acute accent should be dropped. Y/N
 * 4) The dieresis accent should be dropped. Y/N
 * 5) The letters æ and &#339; should be represented respectively as ae and oe. Y/N

I believe we have consensus on point 1 and 2. What of the other points? How far do we go in towards extreme Anglicization. I currently prefer, as does Wiglaf, I believe, that the answer should be no to points 3–4 and that we standardize on the more technical usage found most recently in Cassel's and in John Lindow's writing (and in older usage in Brodeur among others). But is there consensus? Jallan 04:03, 25 Jul 2004 (UTC)


 * Since this page is running over 35 kb and is getting rather long, I have taken the liberty of shortening it considerably. I hope I don't offend anyone by doing so.


 * I think we're arriving at something here. Thanks for the interesting review!

1. Yes. The letters þ, ð and &#491; should be represented respectively by th, d, and ö (not considering in this point whether ö should be retained as ö or changed to o which is the subject of point 4). 2. Yes, the dropping of case endings should be acceptable to most people. I don't understand why a case ending should be retained in a non-case language, even if it is in the nominative. With exceptions for certain well-known and monosyllabic names, such as Baldr, Hödr and perhaps Freyr. For pronunciation reasons I suggest "Balder" and "Höder" to be the exceptions to this rule. Moreover, "Balder" is the most common English and Scandinavian form, and there's no reason to exclude his brother Höder from this, as they most famously appear in the same story.
 * For what it is worth, in their definitions on Balder the Oxford English Dictionary gives the name as Hödur but Webster and Britannica both give Höd. Höd is even a short stub entry in Britannica. :-(  Jallan 21:49, 25 Jul 2004 (UTC)

3. I hesitate. Odin and Thor are household forms in English language popular culture. The accute accent should be noted in the articles, though.
 * If acuate accents were generally represented, Odin and Thor would come under the rule of keeping forms that had become common irrespective of what rules are otherwise applied. For that reason Brodeur and Cassel's also leave the acute accent off these names though applying acute accents otherwise. Jallan 21:49, 25 Jul 2004 (UTC)

4. No. Ö is too well-known in popular culture to be dropped.It also has Nordic and Germanic connotations.
 * The English dictionaries I consulted don't provide any entries for an Old Norse ö name that I could discover. But as mentioned above I did find Hödur and Höd in entries on Balder, which suggests a similar feeling that one should retain ö at least. Jallan 21:49, 25 Jul 2004 (UTC)

5. Yes, I suggest that for typing reasons, the letter æ should be represented as ae. &#339; should be represented by Ö.
 * I'm not sure that typing reasons are relevant. For people using the US Windows keyboard ö is just as much a problem as æ is. But the æ digraph is not especially common and in any case, whatever is decided, there should be redirects for alternative forms. That is, even if we use Æsir and Ægir, a user should be able to get to the entry by typing aesir and aegir respectively. One should also be able to ignore diacritics in typing. For example, I just tried to reach the article Chrétien de Troyes by entering Chretien de Troyes and it worked, because someone had created the expected redirect. The æ question also effects Old English names, indeed much more so as it is more common in Old English than in Old Norse. Some histories and translations have forms like Ælle and Æthelfrith and some use Aelle and Aethelfrith. The Wikipedia entry Aethelfrith has two occurrence of Æthelfrith in the text, one being the first mention of the name. The original article had Æthelfrith  throughout but editors have been, perhaps unwittingly, changing forms. The article called Aelle of Sussex uses the Ælle form throughout. One would expect sources to be consistant on using æ or ae, that is, a text using the form Ælle would also use Æsir and a source using Aelle would use Aesir.  However the Columbia Encyclopedia uses æ in Old English names as in Ælle at Sussex, Kingdom of but uses ae in Aesir in Germanic religion. Jallan 21:49, 25 Jul 2004 (UTC)
 * I'd like to add:
 * 6) The Scandinavian Å is to be dropped, if it appears, as it neither existed in Old Norse nor exists in English. It does appear in modernised Scandinavian forms, but should not in English Wikipedia.
 * That leter å doesn't appear in standard Old Norse or any Anglicizations of medieval Scandinavian forms that I've seen. So that fortunately is not an issue. :-)  Jallan 21:49, 25 Jul 2004 (UTC)


 * It would be interesting to know what others have to say about these points.-- Wiglaf 09:01, 25 Jul 2004 (UTC)
 * I think I will drop out of the discussion for now. The discussion about spelling in the main body of articles isn't really relevant to what I am going to do, and consensus should be reached by speakers of the language in question. I did, however, edit Freyja according to Old Norse convention to see how it looked and which reaction it would get. Feel free to edit back. :-) Cheers Io 11:57, 25 Jul 2004 (UTC)
 * I have no personal problems with this editing which is pure Brodeur. I'd personally favour just removing the case endings, as you also suggest. But one can't impose one's own personal tastes here. Unfortunately once I began to investigate carefully rather than rely on impressions I found that extreme Anglicization of Old Norse names was far more common than I had realized, and that more often than I would have thought extreme Anglicization of Old Norse name was done by choice rather than forced by typographical constraints. (But the expense of using special characters and fancy typography increased exponentially throughout most of the twentieth century and that probably played a part in forming this aesthetic, or as the 1913 Webster Dictionary gave the spelling, æsthetic. Even less than a century ago use of æ was the correct spelling in standard American English for ae in most Latin-derived words. Only the invention and spread of TeX in the middle 1980s began to counteract the increasing inability of printing systems to handle an extended character set or typographical complexity and only within about the last three years have people begun to be confident about exchanging characters electronically that were not part of invariant ASCII. Even two years ago people would have been more leary about exchanging æ or any accented characters because some browers might not support them properly.)
 * The current Webster's College Dictionary gives Aesir as the spelling, as does the Collins Dictionary. However the Oxford English Dictionary (2003) has the form Æsir. Oxford also explains in other entries that Balder was god of the midsummer sun (a somewhat outdated POV theory) that Thor was son of Odin by Frigg, and that Freyja was goddess of the night. There is garbage everywhere. :-(
 * Jallan 21:49, 25 Jul 2004 (UTC)

OK, what about the following standard:
 * 1) The letters þ, ð and &#491; should be represented respectively by th, d, and ö.
 * 2) Nominative case endings should normally be dropped (except for certain one-syllable names. Since the most common one is Baldr, "Balder" is suggested to be the standard. For consistency this would apply to his brother Hödr as well.)
 * 3) The acute accent should not be dropped unless we are dealing with common forms in the English language, such as Odin and Thor.
 * 4) The dieresis accent should be kept.
 * 5) The letter æ should be retained. The letter oe could be rendered as Ö.--Wiglaf 14:12, 27 Jul 2004 (UTC)

Yet more information
I am personally mostly agreed you, Wiglaf, other than on the matter of Höder and &#339;. Höder doesn't appear much used as far as I can see, Balder being now an isolated occurrences of what seems to have been a long obsoleted custom of adding e before an -r nominative case ending. Cassell's gives Höd for example. As to &#339;, the name Hoenir is fortunately unusual enough under any spelling that forms can be searched in Google. I've done this both with Hoenir and Aesir, both with Google set for any language and for English only. Results are ordered by English hits.

Search pattern              Any language  English only H&#339;nir                                218           111 Haenir                               781           607 Honir                              1,650           650 Hænir                              1,070           689 Hoenir                             2,280         1,670 Hönir                              3,640         1,790 Aesir                             47,600        34,100 Æsir                              49,100        38,700

Unfortunately a look at the entries found indicates oddities in Googles identifications of vowels. A search using æ finds also any words using ae but the opposite is not true. Similarly a search using ö find any words using oe or o, a search using o also finds ö, but a search using oe finds neither ö or o.

I repeated the search using the negativising hyphen to remove false matchines, though articles giving more than one form will also be dropped. Results are again sorted by English usage.

Search pattern              Any language  English only Hönir -Hoenir -Honir -H&#339;nir        1,150            63 Hænir -Haenir                        285            82 H&#339;nir -Hoenir -Hönir -Honir          190            90 Honir -Hoenir -Hönir -H&#339;nir        1,490           213 (In place of buggy result 3,570) Haenir -Hænir                        765           603 Hoenir -Hönir -Honir -H&#339;nir        2,120         1,590 Æsir -Aesir                        5,700         4,480 Aesir -Æsir                       42,400        32,800

For some reason when searching on Honir -Hoenir -Hönir -Hœnir an obviously wrong figure occurs. Checking the actual pages of entries returned indicates the true tally is 213.

The standard internet forms in English are definitely Aesir and Hoenir. Use of the form Hönir is indeed very common, but only outside English according to the results. I imagine it is used especially in modernized forms in some Scandinavian languages and in German, which is quite understandable. But it is not at all the common form in English usage.

The net is somewhat conservative in respect to special characters compared to printing which partly explains low usage of æ in Æsir and even more so low usage of &#339; in H&#339;nir as that character does not even occur in the Latin-1 character set. But the results do accurately indicate that US-ASCII forms are more common in Engilsh than forms which represent the original Old Norse more closely. However common use is not necessarily best encyclpedic use, othwise Cassel's and John Lindow and others would not also be providing more accurate forms of the names, especially in encyclopedia/dictionary kinds of works.

I recommend waiting for at least 30 days for possible further input. There haven't been enough participants so far to make clear that there is group consensus on Wikipedia (by those who care at all one way or the other). But if no opposition arises before September 1, I would think it then fair to consider the matter decided, that people other agree or don't care one way or there other. We should then (or preferably before) draw up a short explanatory Wikipedia standard recommending use of diacritics and digraphs for Old Norse names giving reasons and should also recommend the use of digraphs as standard in Old English names. Then one can begin standardizing articles with a clear conscience.

The Cassell's rules, which I cited, seem simple enough and can be adopted. But usage does not have to be entirely consistant between articles. I would also be willing to go with ultra-English if a majority appeared to push it, but if no-one cares cares one way or the or people are tacitly happy with what we have tentatively agreed to, then I think it reasonable to take that as a mandate.

Jallan 02:54, 28 Jul 2004 (UTC)

I completely agree that we'd better wait until september 1st, before establishing some kind of standard. Two people aren't enough to establish a consensus at Wikipedia :).--Wiglaf 10:07, 28 Jul 2004 (UTC)


 * I'll add a third voice here, then. :) I believe that since we're in an English Wikipedia, if an English spelling exists we should use it. Admittedly, redirects are wondrous things, but I just popped up at a page on the Alþingi and if I weren't a linguist, I'd have no way of recognising the word as Althing, the Icelandic parliament. Thorn and eth are no longer letters that the average person is familiar with the pronunciation of, so I think while we should retain the proper Norse-style spelling within the body of the article, the titles of the pages should be Anglicised. thefamouseccles 04:47, 03 Oct 2004 (UTC)


 * I agree with thefamouseccles. If an English version of the name exists, it should be the primary name on the English Wikipedia, with the non-English forms as redirects. Jonathunder 18:17, 2005 September 8 (UTC)

normalization
I like Jallan's normalization effort. Consistency will of course be almost impossible to achieve, since users will add stuff all the time without bothering to search for conventions first, but it is pleasant to even have such conventions. One point though, I find it strange to preserve length marks (acutes), but to drop thorn and eth, e.g.:
 * Hrafnagaldur Ódins (Hrafnagaldur Óðins) 'Odins's Raven Song'

I think this could as well be
 * Hrafnagaldur Odins (Hrafnagaldur Óðins) 'Odins's Raven Song'

since we give the precise spelling in italics anyway. i.e. I think both Óðins and Odins are fine, but Ódinsseems a bit strange. A minor point of course, and not really an objection. dab (&#5839;) 09:57, 2 Jan 2005 (UTC)


 * I tend to agree. I'd rather have diacritic marks AND thorns and eths or no such frills at all.
 * My suggestion at the moment is to have a hypercorrect naming such as Óðin inside texts, but Odin as the name of the article. --Wiglaf 19:18, 2 Jan 2005 (UTC)

Response to initial comments
Reduction of Norse names to the 26-letter alphabet with no diacritics is common in English. I will go along with that, if that is indeed what other people really want here.

Personally I don't prefer such lowest common denominator renderings.

The opposite idea, using the full Old Norse forms throughout is just not English practice. Rudolf Simek's A Dictionary of Norse Mythology (translated from German) is the most extreme example I am aware of in the direction of using Old Norse forms without change in a work on Norse mythology. That work uses the hooked-o, and &eth; throughout, still retains case endings on names, and always renders thorn as th where it occurs. But even this text still uses the English spellings Odin and Thor. Only some scholarly articles in international journals are more or equally extreme. Jean Young's partial translation of Snorri's Edda also used &eth;, but dropped the case endings and used ö rather than hooked-o. These works are the unsual exceptions. Normally, in translation after translation, commentary after commentary, you get &eth; rendered d or more rarely th or occasionally dh, diacritics more often dropped than included, and æ sometimes retained and sometimes rendered ae, and various things done with &oelig; when it is not retained as is. I could personally accept following Simek, except that I don't think that such usage fits the general Wikipedia principle of using the form of a name most common in English. Which forms are most common in English for names from Norse mythology is very debatable, but they are certainly, and perhaps unfortunately, not the Old Norse forms with case endings and all the "frills". And in general the policy of using the most common names serves Wikipedia well, even though I'd love to start using more correct forms like M&#333;&#353;eh or M&#333;sheh instead of Moses and Akhilleus instead of Achilles. I tend to be a pedantic purist. But the current policy rightly keeps pedantic purists from being able to replace accustomed English forms with more correct forms, at least until other works start doing the same thing. Don't come here to change the standards, even if the change would be for the better. And it keeps those who want to use non-standard systems from fighting over which non-standard system. Unfortunately there is no standard system for Old Norse, but there are features of adapting the names that are more common than others.

I was using the forms used by Lindow and Orchard because they represent the most normal English letter spellings (j not i, d not th or dh for &eth;) and their rules mostly provide naturally the forms found normally in English, for example, Sigurd, not Sigurth or Sigurð. The diacritics are also found in many translations and are the kind of extra one expects in more scholarly works. Brodeur used the same forms, but kept the case endings. Hollander used the same forms except for rendering &eth; as th and ö as o; but he kept the other diacritics. Faulkes and Larrington use the same forms, but drop the diacritics and represent j by i. Lindow and Orchard represent a reasonable common denominator rendering which is yet not a lowest common denominator rendering. They keep æ rather than using ae and keep &oelig; rather than rendering it as oe. They keep the diacritics, which readers can mentally filter off, if they wish, but can't filter on again if they aren't there. And they provide an external standard that can be followed without niggling over whether a name is well-known enough in an English form or unaccented form to appear that way. Óláf or Olaf? Just see what they do and copy it. And because they have written encyclopedic works, they have far more forms than those found in most individual translations. I could compile a list of the name which are exceptions to their normal rules for the Manual of Style. There aren't many: Odin, Thor, Halfdan, Olaf are the only ones that spring to mind. Orchard only identifies Odin and Thor and just mentions there are others where he has used forms "more familiar to the general reader". When discussing Old Norse words as technical terms, both Orchard and Lindow do use accurate Old Norse forms.

I don't like the suggestion of different forms as article headers and as standard forms within articles, except for technical reasons. If we adopt a standard form of a name for Norse mythology in Wikipedia, it should normally be the standard form in both the title and article. It seems to me there would be something philosphically wrong with a policy in which almost all links would be piped links. If the common Wikipedia spelling of the Norse king of the gods is decided to be Ó&eth;in, then the article should also be Ó&eth;in, with redirects from the forms O&eth;in, Ó&eth;inn, Odin, Ódinn, Othin, Óthinn, Odhin, and Ódhinn. If we can't justify Óðin as the proper title for an article, then it shouldn't be the normal form in the text either.

I do see reason for making a partial exception in the case of anglicized titles of Norse literary works. In an article they might more often be referred to by their Old Norse form, for some more than others. Voluspa doesn't look too bad. Fra daud Sinfjotli looks like carelessly rendered Old Norse. However Thrymskvida or Thrymskviða seems far better to me than Þrymskviða in a casual reference in English. Such differences is the impression such forms make is probably one reason there is no consistancy in Wikipedia. One could also make an exception that titles of works, but not other names, do not follow the Orchard/Lindow rules, but appear in Old Norse forms following Simek's rules, that is substitute th for thorn but otherwise follow the Old Norse spelling, even to the use of the hooked-o, though currently in article titles ö would have to be used as a substitution. I think it would be simpler just to stick to Orchard and Lindow's reasonable and consistant anglizations. Different forms tend to look strange, pedantic, or dumbed down to different readers.

I have already been doing a fair amount of standardization based on what I believed was consensus. If consensus has changed now, fair enough. I am willing to change what I've done. But consensus should not keep changing. One way is often as good as another and there are thousands of divergent forms now regardless of what standard is used. For search purposes I want to continue providing alternate forms at the bottom of articles, (see Halfdan the Old as example). Doing this partly depends on having standard forms within an article, whatever that standard is. A change of standard away from Lindow and Orchard now will require quite a bit of work on my part to fix up what I've already done. But I'd rather this happen now than six months from now after making many more such lists of variant forms, which is why I am bringing up the matter after producing a concrete example of what I have in mind. I want to get things started, if possible, and I want what I am doing to have support by those who are currently working in this area, otherwise it won't be done. And it should be a standard that others cannot lightly change in the future, if only because it has become, hopefully, so entrenched by it being followed that the work involved in changing it would be enough to put people off.

We're past the point where we can just change a standard lightly and start fresh with new rules every few months. And the mess of having no rules is becoming more and more annoying. We must decide. And we must start using whatever standard is decided on, or watch the anarchy increase, and face the possibility of silly edit wars over spelling. Generally, it is better to be using a standard for names that one does not particularly like than to be forced to fight again and again on article after article to get spellings accepted, which is why such standards exist. It enables one to fight instead about what is worth fighting over, the content. I far prefer Ægir to Aegir (and use of æ does appear in some translations which don't use diacritics.) But if I am outvoted, it is not worth fighting about compared to the time taken to actually improve an article about Ægir or Aegir. But if we don't decide, we are making more and more work for ourselves in the future.

Jallan 01:02, 3 Jan 2005 (UTC)

There is no perfect solution - here's what I do and why
Personally I prefer the following standard:

1. Using no characters outside the common ISO-8859-1 (a.k.a. Latin-1). Instead of o-with-tail we can use 'ö'. Instead of 'oe-ligature' we can use 'oe'. Instead of 'o-with-tail-and-accent' we can use 'á' (representing a slightly later stage in the language).

I think that using characters outside of Latin-1 is just raising the bar unnecessarily high. I can't read those letters on my personal computer system nor on other systems I habitually use.

2. Using all other ON characters, including 'þ', 'ð' and 'æ'.

These characters, on the other hand, are readable by a large part of the computer systems of the world.

3. Retaining nominative endings.

One argument for this, as well as point 2, is to maintain consistency with Modern Icelandic within Wikipedia. I don't think many would argue that a modern person named, say, "Freyr Þorláksson", should get an entry under "Frey Thorlaksson" on the English Wikipedia.

Examples:


 * Ørlög
 * Hallfrøðr
 * Freyr
 * Höðr
 * Þorbjörn
 * Óðinn
 * Hoenir
 * Öku-Þórr
 * Alþing
 * Áss inn hvíti
 * Æsir
 * Hildr

This is what I have done in the articles I have written. Here's a work in progress: Hrafnkels_saga

It is also the standard I apply in my Old Norse for Beginners course:

http://www.hi.is/~haukurth/norse/

I originally wrote Old_Norse_orthography to help clear the matter up. Ultimately the only possibility for beginners to make sense of the various possibilities is to read articles like that.

I'd like to clear up a misunderstanding I see in some of what's written above. The 'r' in Baldr is not a nominative ending. The accusative is Baldr, not **Bald. The accusative of Höðr is Höð, on the other hand.

Haukurth 01:04, 20 Jan 2005 (UTC) (edited a few minutes later)
 * Are you aware of the   template? This allows you to specify that characters or words should be rendered in a unicode-capable font. An example from above might be "Njörd (Old Norse Nj&#491;rðr)" which now renders fine on my browser. (The template specifies freely-available fonts BTW so we're not asking people to spend money.) --Phil | Talk 08:34, Jan 20, 2005 (UTC)


 * That actually seems to have worked :) I'm seeing some &#491;'s on my screen now. I'm guessing this is not possible to employ in page titles? Now I'm wondering if I should switch to using this letter (as well as oe-ligature). When listing my sources in the Hrafnkels_saga article I had a book with an &#491; in its title. Normally I would have transcribed it with 'ö' (as would be done in any Icelandic library system) but now I tried using the o-with-tail character. Haukurth 18:03, 20 Jan 2005 (UTC)

No Agreement
Well, obviously there is no agreement among any of those commenting here. Accordingly there is no point in my doing any further standardization work. I will note that I cannot offhand think of any English source which uses modern Icelandic spellings of Old Norse names. They either use Old Norse forms more or less Anglicized or Anglicized forms based on modern Icelandic forms.


 * If you come to Iceland as a tourist you can find texts on "Ingólfur Arnarson" in English. I'm not suggesting we do that, though. It may not be that easy to draw the line, however. What about a 15th century Icelander? How do we spell his name here? - Haukur

I have not been able to locate any style guide indicating how modern Icelandic names should appear in Modern English. One sees them also in various formats, especially with &thorn; rendered th and often with &eth; rendered as d. I don't know to what extent such uses are based on actual style guides and to what extent they are ad hoc. For example, after setting Google to English, I find 82 hits for ["&THORN;orger&eth;ur Katrín Gunnarsdóttir"], 26 hits for "Thorgerdur Katrín Gunnarsdóttir" -"Thorgerdur Katrin Gunnarsdottir"], and 27 hits for ["Thorgerdur Katrin Gunnarsdottir" -"Thorgerdur Katrín Gunnarsdóttir"]. English translations without &thorn; and &eth; and without diacritics are used on the Icelandic govenment website. See http://government.is. But some other pages on the same site use Icelandic spellings.


 * Indeed. This is unfortunate. Various systems are in use and it is often difficult to make people even understand the problem. Most seem to think that the way they do it is the obvious and only way to do it. - Haukur

The difficulty in arguing that we should follow modern Icelandic use, is that most English translations of sagas and other Norse texts do not so. That is, such translations indeed normally use Frey, but would, if referring to a modern Icelander named Freyr in an introduction or bibliography, indeed use the form Freyr in that case. The inconsistancy is commonly recognized. Also &oelig; is replaced by æ in modern Icelandic and nominative -r by -ur in most cases. Accordingly &THORN;orger&eth;r in modern Icelandic is &THORN;orger&eth;ur and Baldr is Baldur.

Some English translations and commentaries do favor modern Icelandic forms of Old Norse forms. Most do not, or if they do so, do so, use them without case endings.

Haukurth's method seems to me to be yet another idiosyncratic rendering system, but as it is used, as far as I know, in no published texts, it should not be the basis of a standard in Wikipedia. It is an original research transliteration system.


 * Not quite. Icelanders very commonly substitute ö for o-with-tail. Try searching for "Austfirðinga sögur" in a computerized library system. See if "sögur" is spelled with ö or with o-with-tail like in the actual publication. See www.hi.is/~eybjorn for another example. He may use the oe-ligature character, however. I also use this system outside of Wikipedia and &lt;tongue-in-cheek&gt; since my website is the number one Google hit for "Old Norse" it is obviously the de-facto standard&lt;/tongue-in-cheek&gt;. - Haukur

As to raising the bar too high in using special characters, the &oelig; character, at least when entered as &amp;oelig;, should be available under 99% of browsers in current use (and those that won't show it also won't properly display true dashes, such as those used throughout the Main Page).


 * When I'm working on Linux I often don't see the oe-ligature. But the problem isn't limited to seeing or not seeing. Now I'm working under a Microsoft operating system. When I copy text including an oe-ligature from Wikipedia and paste it into Notepad (surely not an unreasonable thing to do) I get a black box where the oe-ligature letter was. I like being able to copy text without having to hand-modify it afterwards. - Haukur

Limitation to Latin-1 is very much not a Wikipedia policy. There are thousands and thousands of English Wikipedia articles using Chinese/Japanese/Korean characters, International Phonetic Alphabet characters, Hebrew characters, Arabic characters, special Latin letter transliteration characters and so forth. Some of these articles have been featured articles.


 * Of course. But that's inevitable. We can get around it in Old Norse. - Haukur

The o-hook character is more arguable (because not available in most Latin letter fonts), and in any case is never used in Anglicized versions of Old Norse characters (and not used in modern Icelandic). Generally it appears in English works in italicized forms only, the italics indicating that this is a genuine Old Norse form as opposed to an Anglicized form. (The main exception is in scholarly journals.) To produce o-hook with an acute accent the proper Unicode method is &#491;&#769;. This may look clumsy on your browser.


 * Yes, or it may not show up at all. Like on mine. Now. Under Internet Explorer. - Haukur

(On my browser, using the font it chose, the acute is far too high.) In theory the acute accent, by Unicode rules, should be positioned properly over the base character. In fact most fonts do not support variable positioning of combining characters, so the result may look poor, unless your browser happens to choose a font which by default positions the combining acute in the proper position or chooses an intelligent font that automatically adjusts the position of the acute to fit the base character. But this is the proper Unicode method. Unicode's policy from the beginning was that using combining diacritics was preferrable to using precombined characters which were included for compatibility reasons. But font technology has lagged behind. However Unicode's policy is now to include no more characters which represent a combination of a base character and one or more diacritics, in part because there is no end to such combinations.

In any case, I now do not expect to be active in Wikipedia for some time (if ever), and accordingly would not have been able to get on with immediate standardization in any case.

Jallan 18:30, 23 Jan 2005 (UTC)

Okay. I think we can agree that there is no agreement. I could personally live with Anglicized names used throughout the text if the full form of the Old Norse name is given at each name's first occurrence. So, what's your favorite standard for Anglicizing names? I don't see any agreement on that. How do you, for example, handle oe-ligature? It would seem logical to replace it with "oe" but I'm not sure that's widely done. And should we drop nominative-r? Should I spell "Sámr" as "Sam" in the Hrafnkels_saga article? It seems silly and unnecessary but if that's what people want I'll go along with it in the name of having a standard. Haukurth 20:52, 23 Jan 2005 (UTC)

I had forgotten that Jallan had already done some research on replacements for oe-ligature. Apparently the most common thing on the web is to replace it with "oe" (which is what I was originally advocating).

I guess this whole thing is somewhat more personal for me than some of you. I actually have a thorn in my patronym (which is not a surname). My full name is Haukur Þorgeirsson. When I'm forced to replace the thorn with something I choose 'th' (like in my user-name). I do this consistently but that doesn't mean I like it. Imagine if Icelandic was a standard computer language and English wasn't. We might not bother with silly "frills" we don't use like the letter 'w', which after all is not a part of the original Latin alphabet. Imagine Mr. Wilson having to write his name VVilson (and then arguing with other VVikipædians on the relative merits of 'VVilson' vs. 'Vvilson' vs. 'UUlson' etc.). Haukurth 23:24, 23 Jan 2005 (UTC)

I've been experimenting a bit in the sandbox and it doesn't seem possible to get oe-ligature or o-with-tail to show up properly in page titles. Is there a way around this, is it a problem with my system or do we have to avoid these characters in page titles? Haukurth 10:11, 24 Jan 2005 (UTC)

I'm editing some mythological articles now and generally I use the standardized ON spelling for the names. I make an exception by using "Odin" and "Thor" instead of "Óðinn" and "Þórr" since I think these names are actually familiar to English speakers.

I still don't know if it's possible to create page titles with o-with-tail and oe-ligature.

Does anyone out there want to talk about this? Haukurth 12:35, 18 May 2005 (UTC)


 * Well, yes, but I wonder whether it is possible to arrive at a naming convention.--Wiglaf 12:39, 18 May 2005 (UTC)

Category:Wikipedia naming conventions

No original research
It seems to me that some of the things talked about on this page are in danger of breaking the WP:NOR. Perhaps a constructive way forward is to list some of the options which have been used by various people who have translated and published a lot of texts for the general public to read (not specialist to specialist) and comment on how they have translated the names etc. It there is one person, or School of people, which has a substantial body of work that covers a wide area, and it is a close fit to the Naming convention then perhaps that could be used as a standard.

Even if nothing else comes of such a survey it it would at least produce enough information to help make a more informed choice on how Wikipedia editors could progress the issue and probably produce enough information for a page on the history of translations. Philip Baird Shearer 17:28, 7 October 2005 (UTC)

WP:NOR is surely not intended to prohibit Wikipedia from deciding its own naming conventions. In any case there's nothing wrong with following an external standard. That's basically what we're doing. We've taken the spelling used in works like Ursula Dronke's translation of the Poetic Edda (the best English translation there is, no competition) and Richard North's recent book on Haustlöng (an excellent and very accessible work) and made some minor adaptations. Those are the use of 'ö' instead of o-with-hook and the use of familiar English forms where those exist (like Odin and Thor) in accordance with Naming conventions (use English).

Of course other external standards are possible. Jallan proposed the use of the forms in John Lindow's Handbook of Norse Mythology. Those can be seen at User:Haukurth/Handbook. This proved somewhat unpopular because, well. Let's take an example:

We have a Lindow name like "Ögmundar tháttr dytts ok Gunnars Helmings" and Jallan's proposal was to use that in the title. But we also wanted the full ON form with 'þ' so we'd list that too right after the other:

Ögmundar tháttr dytts ok Gunnars Helmings (Old Norse: Ögmundar þáttr dytts ok Gunnars Helmings)

But that looks somewhat confusing. Readers are unlikely to realize that the first version (which includes such characters as 'ö' and 'á') is actually the anglicized form. They'll be likely to incorrectly assume that a spelling which includes 'ö' and 'á' is the original spelling.

The basic problem is that just about every translator or editor uses her own spelling system. I could stay up all night showing you the myriad of systems in use. The only really standard forms are the un-anglicized Old Norse versions, as used in any number of English works. Many of them are scholarly, granted, but then many of us feel that Wikipedia should be somewhat scholarly as well so we're not troubled by that. Of course we should make allowance for Odin and Thor, use redirects liberally, list alternative anglicizations (see Lóðurr) and generally try to please everyone. Please tell us how we can please you. I know you want to see alternative anglicized forms and ascii forms in the lead and I agree with that. How about sound files with the pronunciation of those tough names? IPA transcriptions? A more organized effort to create redirects? - Haukur Þorgeirsson 18:02, 7 October 2005 (UTC)

THIS IS AN ARCHIVE. PLEASE DO NOT POST HERE. GO TO Wikipedia talk:Naming conventions (Old Norse/Old Icelandic/Old English) INSTEAD.