Wiktionary talk:About Middle High German

Questions
There is (a) the old, real, natural, chaotic MHG as spoken by Germans in the Middle Ages and (b) the artifical, organised MHG as invented later. E.g. compare the texts at de.wikisource.org/wiki/Dû_bist_mîn,_ich_bin_dîn with the Middle Age manuscript. The wikisource texts have "sluzzelîn" (modern) resp. "sluzzelin" (pseudo-historical), while it is like "ſluZZellın" (long s, capital Z or a letter similar to it, i without dot) in the manuscript. -eXplodit (talk) 17:06, 25 July 2015 (UTC)
 * 1) Are the two Middle High German language variants somehow separated?
 * 1) Which letters are or shall be used?


 * I guess there are much more MHG language variants than just two: they had dialects, too, and there was a difference between the language of the common man and the literary language. See for instance Höfisch. Sociolects may have existed, too.
 * Grammars and dictionaries usually normalise, not only MHG ones but also Old Norse, Old High German, Middle Dutch ones. We could check our MHG grammars and dictionaries to see how they write: v- vs. f-, ī vs. î, z vs. ʒ and so on.
 * I think we should normalise (as we do with Old Norse terms). Otherwise we will have a large amount of variants. And since there were no language or spelling rules in the Middle Ages, every typo could be a legal variant.
 * --MaEr (talk) 17:25, 25 July 2015 (UTC)
 * --MaEr (talk) 17:25, 25 July 2015 (UTC)


 * Yeah, there were dialects. But I'd guess that that is less problematic as English, German, (Classical) Greek and many other languages have dialects too.
 * About Old Norse states that attestable real spellings can be mentioned too, so I don't see any problem in having normalised spellings as the main lemma.
 * Regarding normalised spellings:
 * In zit it is "zīt" (with a macron), and Wright's Middle High German primer uses macrons too (e.g. "mīn" and "dīn"), and duden.de and dwds.de use macrons too (duden.de: Zeit, DWDS: Zeit). Though German MHG dictionaries (woerterbuchnetz.de include some) seem to prefer circumflexes as in "zît". Well, maybe macrons are the better variant, as they are also used in Latin and Greek and might be more intuitive.
 * Wright's Grammar uses ȥ, duden.de and dwds.de use ʒ (e.g. in the etymology of Duden: essen, DWDS: essen), and German MHG dictionaries use ʒ too. German MHG grammars have both forms, though it might be that ȥ is rather used in older ones, while it might rather be ʒ in newer ones.
 * waʒʒer/waȥȥer was requested as wazzer in Requested entries (German) - and the distinction between z and ʒ/ȥ might be modern/artificial. So should it be like this?
 * Regarding templates and case orderings: The noun declension template here (Template:gmh-decl-noun) has "Nom. - Acc. - Gen. - Dat." which is also used in Wright's grammar. German MHG grammars have "Nom. - Gen. - Dat. - Acc." which is also used in German. Having MHG "tac - tac - tages - tage" while it is NHG "Tag - Tages - Tage - Tag" is kind of irritating. So, how about using "Nom. - Gen. - Dat. - Acc." to have it more similar to German entries and to make it less confusing?
 * -eXplodit (talk) 00:16&00:29, 26 July 2015 (UTC)
 * I never understood the practice of separating nominative and accusative like that in IE languages. They are often the same, and they are always the same for neuter forms (a property inherited from PIE). So I prefer to put accusative after nominative in MHG and modern German. —CodeCat 00:30, 26 July 2015 (UTC)

PS: Well, sortable templates aren't a (good) solution (example follows hereafter in the source code). But maybe there are other "tricks" (depending on javascript) to inlude two tables while only showing one.
 * Traditionally cases were ordered "Nom. - Gen. - Dat. - Acc. - Voc. - Abl." (regarding Greek without abl.), so it simply is the traditional order. In case of German, the cases also have other names like "erster Fall (first case) - zweiter Fall (second case) - ...", so re-arranging the cases would lead to confusion.
 * IMHO those who re-arrange cases and genders, often use illogical systems (too). E.g. in NHG they (just "usually"?) use "m. - n. - f. - pl.", more logical would be "n. - m. - f. - pl.", as the masculine form is sometimes similar to the neuter and sometimes to the feminine.
 * It's easier to change the MHG template as there are only a few MHG entries. To change the NHG templates and entries would be much work (and maybe users would oppose such a change).
 * Maybe one could "somehow" use variable templates. Maybe like using sortable templates, or using two templates and some other "tricks" so that users can choose the template they want to see.


 * -eXplodit (talk) 01:04, 26 July 2015 (UTC), PS: -09:41, 26 July 2015 (UTC)


 * Regarding which letter-shapes to use: per longstanding consensus, don't use long s anywhere (except, optionally, citations) in any language; the software automatically redirects anyone who searches for a long-s entry to the right place. Likewise, don't use dotless i except in languages like Turkish where it contrasts with dotted i. (And don't take notion to misuse ɑ to represent an upper-bar-less a, or misuse uͤ or ű to represent different shapes of ü, etc.) I agree with MaEr re normalization of spelling. We could, of course, have pointers from the various manuscript spellings to the normalized spelling, like hliod, except that we should probably devise a dedicated template like manuscript spelling of rather that using altform with a context label. - -sche (discuss) 03:00, 26 July 2015 (UTC)

Well, one problem might arise from mentioning spellings with long s: "Earlier" there were no rules or no single rules regarding the use of long s, so one might find old spellings (16-18th century) which contradict the modern spelling rules (19-20th century, esp. after 1902). But with qualifiers indicating the time etc. this shouldn't be a real problem. manuscript spelling of is fine, but shouldn't this also be used in case of Old Norse?
 * Letter shapes: Mostly I can understand that and I'm mostly okay with it, but not completely:
 * Long s: I can understand that long s shall not be used in page titles and headers. But at least sometimes it should be mentioned under "Alternative forms" (without a link of course) like it was for example done in Gasthaus, as there isn't or isn't always a bijection between spellings without long s and spellings with long s (e.g. compare Wachstube with Wachstube/Wachſtube). Also: Long s isn't forbidden (at least not forbidden anymore) or (completely) deprecated in German, and the "Bund für deutsche Schrift" (BfdS) for example still uses fraktur and long s.
 * Umlauts: There's a bijection between the spellings, so there's no problem in just using the modern forms. But IMHO it would be nice to mention that there were different forms on the about pages. But okay, maybe mentioning alternative letter forms could be too long (ʒ vs. z for the ts sound, ligatures), and in a way it is unnecessary.
 * Normalised spellings: About Old Norse says that there can be entries for real spellings (if they are attestable), and same should be done in case of MHG.
 * -eXplodit (talk) 09:41, 26 July 2015 (UTC)


 * Re the zs: I would follow the Duden and de.Wikt and prefer ʒ to ȥ if we decide to use one of those when normalizing. I note that 'pedia no special letter was used in the manuscripts (only straightforward s, z, ss, sz or ß). - -sche (discuss) 21:22, 26 July 2015 (UTC)
 * I think we shouldn't use either of those letters, we should just use z. That was the letter used in the manuscripts; how it looked is not relevant. —CodeCat 21:28, 26 July 2015 (UTC)

When we differ between z and ʒ/ȥ, the questions arises whether article titles should be like (a) [[waʒʒer]] or [[waȥȥer]] or (b) simply [[wazzer]]. I do prefer (b) as people usually can't type ʒ/ȥ (me included, I copy&paste them), and as the distinction between z and ʒ/ȥ should be a modern invention (e.g. in Nibelungenlied C p.1 it is "grozer" [in "grozer arebeit" and "tugende. grozer eren vil gewann"], while Wright's grammar implies the spelling "grōȥer" and while German dictionaries imply the spelling "grôʒer" [German MGH dictionaries] or "grōʒer" [German NHG dictionaries with etymologies]). -eXplodit (talk) 21:40, 26 July 2015 (UTC)
 * @Codecat: In the manuscripts it's z, but normalisied spellings do (sometimes) differ between z (= NHG z, sounding like ts) and ʒ/ȥ (= NHG ß, in IPA s]). And IMHO we should differ between z and ʒ/ȥ. But I'm undecided whether it should be ʒ or ȥ. (My impression is that ȥ is more common in older grammar books, while ʒ is more common in newer grammar books and German dictionaries.)
 * I also think we should use plain z in page names, but we can use ȥ in headword lines, just like we use macrons in Latin and Old English. We can even tell the module to convert ȥ in MHG links directly to z. And it should be ȥ, not ʒ, since the latter is explicitly defined by Unicode as being the IPA character ezh (standing for the voiced postalveolar fricative of "measure"). —Aɴɢʀ (talk) 21:45, 26 July 2015 (UTC)
 * What about OHG? We should really use the same convention for both. And for OHG we currently use z exclusively. —CodeCat 21:49, 26 July 2015 (UTC)
 * @CodeCat: Then it should be better to add the distinction to OHG. duden.de for example writes "mittelhochdeutsch, althochdeutsch grōʒ, ursprünglich = grobkörnig", so it uses macrons and ʒ for OHG too.
 * PS: There were already two MHG entries which use ȥ: fuoz/vuoz; and in fuoz ȥ is also used for OHG. (I doubt that one can call that consesus or common practice, but I thought it should be mentioned.)
 * -eXplodit (talk) 22:03, 26 July 2015 (UTC), PS: 00:31, 27 July 2015 (UTC)

Full orthographic normalisation scheme
Middle High German needs a unified scheme like that given at WT:AGOH to account for various things not mentioned on this page. For example, do we want to use œ and æ for long ö and long ä respectively? Can we (and do we even need to) standardise by choosing one of ʒ or ȥ? I don't have especially strong opinions, but I've been reading Wright's grammar and thought that if at some point I might want to add MHG entries, it would be extremely helpful to know what what to use as the normalised spelling. (and do ping anyone else who might be interested) —Μετάknowledge discuss/deeds 23:25, 28 October 2016 (UTC)
 * 1. We use Ö/Ä with a second diacritic (macron or circumflex) in Middle Low German, but MLG (unlike modern LG and MHG) does not have a tradition of using œ, æ or iu attached. 2. I don't see how we need ʒ. When would there ever be any ambiguity whether ⟨z⟩ is [ts] or [s̟]? 3. We also (and this is due to MLG tradition) use circumflexes for plain long vowels, as do the editions of High German texts that I saw. I assume the price is an inconsistency with OHG/general Wiktionary practice. Korn &#91;kʰũːɘ̃n&#93; (talk) 04:32, 29 October 2016 (UTC)
 * MHG didn't yet have open syllable lengthening, so the only long vowels it had were the inherited ones. Therefore there's no need to distinguish macrons and circumflexes. —CodeCat 13:02, 29 October 2016 (UTC)
 * But we should still choose one. —Μετάknowledge discuss/deeds 18:03, 29 October 2016 (UTC)


 * Actually, I feel this choice has been made for us by Mittelhochdeutsche Begriffsdatenbank, which has lemmatised 10 million words of Middle High German (this is almost an order of magnitude more than the Latin at Perseus) more-or-less using the orthography of the Mattias Lexer dictionary, available here: . Granted that the MHD BDB texts are not public, we would in principle be able to do something like this: User:Isomorphyc/Sandbox3 with properly lemmatised text, provided we standardise around an accepted orthography, at least `within a bijection.'  This mainly suggests:


 * 1) We should use circumflexes, not macrons.


 * 2) We should use ʒ, not ȥ, z, or sz.


 * 3) We should use umlatus, not digraphs.


 * 4) There are a variety of consonant rules which I'm not able to enunciate (for example, older texts use 'v' where newer texts use 'w', and 'v' sometimes for 'f', as for example at the Bibliotheca Augstana version of the Bartsch text of the Nibelungen Leid.)


 * I think these conventions are quite reasonable; the least reasonable is (2), pertaining to ʒ/z. For example, my Deutsch Klassiker der Mittelalters edition of Karl Bartsch's text of the Nibelungen Leid follows (1), (3), comes quite close to (4), but omits (2). (Note the original Bartsch text does not use circuflexes, has different consonants, and is somewhat harder to read.)  I'm not really opposed to using z instead of ʒ, but I do not think the consistency argument with OHG is terribly compelling.  We are not consistent between ModE, ME and OE, and it is easy to paper over the gap with also templates and redirects.  Instead, the reason I would favour 'z' is that it is much easier to type.


 * I'm also sorry this is not terribly well thought out. I haven't worked on MHG in a long time, and I've never given it any thought in Wiktionary outside of occasional modern German etymologies.  Isomorphyc (talk) 02:21, 30 October 2016 (UTC)


 * Thank you! I wasn't aware of that wonderful resource. I think that following those rules would work (well, I don't really understand what you mean by (3), as both umlauts and digraphs are used by Wright and the Begriffsdatenbank, depending on length). As for (2), both ʒ and ȥ are rather annoying for me to type, but we should still represent the distinction (though the pagetitle should have z, of course). —Μετάknowledge discuss/deeds 03:45, 30 October 2016 (UTC)


 * Sorry, I misread your original note. I thought you were saying one would use æ instead of ä, etc.  For what it is worth, the Lexer dictionary has four instances of non-digraph `oe' and five instances of non-digraph `ae'.  I am not sure what the explanation is, but the overlap is so small that either convention can be used, and one will have to write at most nine usage notes. Isomorphyc (talk) 05:05, 30 October 2016 (UTC)


 * Nothing stops us from consistently using ⟨z⟩ in OHG as well, it's equally predictable there. Korn &#91;kʰũːɘ̃n&#93; (talk) 09:45, 30 October 2016 (UTC)


 * There is an extremely small number of contractions and similar situations in which some ambiguity exists, but I would be fine to standardise on 'z'. Isomorphyc (talk) 16:57, 31 October 2016 (UTC)
 * We could, for example, encourage but not require that it be marked on the headword line and when linking, but require that if it is marked, ʒ be the symbol used. —Μετάknowledge discuss/deeds 17:43, 31 October 2016 (UTC)


 * The two best options I can think of are: 1) omitting the curly-tailed z entirely; and 2) substituting automatically, with manual override, per a data module which I can write. A third option is to use an external client (a robot) to add curly-z arguments to head= as necessary.  I would view the third option as ungainly.  Isomorphyc (talk) 14:35, 1 November 2016 (UTC)
 * The second option is the best, in my opinion, if you would be willing to write said module. —Μετάknowledge discuss/deeds 17:17, 1 November 2016 (UTC)

Cutoff Date
@Jberkel What is the cutoff date for MHG and Modern German? Vininn126 (talk) 17:05, 9 October 2022 (UTC)
 * No idea! But sounds like something which should be documented. –Jberkel 19:40, 9 October 2022 (UTC)
 * 1500 per ISO.
 * Alternatively, the New High German vowel shift could be used which occured at different times in different regions, e.g. from /i:/ /y:/ /u:/ to /aɪ̯/ /ɔɪ̯/ /aʊ̯/ as in –  and  – . --08:30, 7 April 2023 (UTC)