Talk:High German consonant shift

From Wikipedia, the free encyclopedia
Jump to: navigation, search
Former good article High German consonant shift was one of the Language and literature good articles, but it has been removed from the list. There are suggestions below for improving the article to meet the good article criteria. Once these issues have been addressed, the article can be renominated. Editors may also seek a reassessment of the decision if they believe there was a mistake.
Article milestones
Date Process Result
July 4, 2006 Good article nominee Listed
October 7, 2009 Good article reassessment Delisted
Current status: Delisted good article
WikiProject Languages (Rated B-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Languages, a collaborative effort to improve the coverage of standardized, informative and easy-to-use resources about languages on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
B-Class article B  This article has been rated as B-Class on the project's quality scale.
 Mid  This article has been rated as Mid-importance on the project's importance scale.
WikiProject Germany (Rated Start-class, Mid-importance)
WikiProject icon This article is within the scope of WikiProject Germany, a collaborative effort to improve the coverage of Germany on Wikipedia. If you would like to participate, please visit the project page, where you can join the discussion and see a list of open tasks.
Start-Class article Start  This article has been rated as Start-Class on the project's quality scale.
Checklist icon
 Mid  This article has been rated as Mid-importance on the project's importance scale.

Shouldn't this page be merged with Second Germanic sound shift? Maartenvdbent 20:46, 7 Jun 2005 (UTC)

Done. --Doric Loon 12:37, 11 Jun 2005 (UTC)

Isogloss movement in the last 15 years[edit]

Since German reunification, a northward movement of the eastern end of the Benrath line has been observed.

This is a remarkable claim. What is the source for it? --Angr/tɔk tə mi 13:19, 11 Jun 2005 (UTC)

Ah. It was in the article on the Benrath line and from there I followed it back to a reference in the German Wiki. I meant to check it further but haven't done so yet. I decided to mention it because it really is interesting, but I did have some reservations. Is your instinct to delete it? --Doric Loon 20:42, 11 Jun 2005 (UTC)

Just added a reference for this claim, however 15 years is not correct. The "extinction" of Low German in most of Brandenburg (mostly in favour of Berlin dialect) and northern Sachsen-Anhalt (mostly in favour of Saxon dialects) began after WW2 and took several decades. Another correction I made is the description of the Benrath line, it's not north of the Uerdingen line all the way but swaps sides with it: Berlin dialect says "ik" and "machen". Anorak2 18:49, 5 October 2007 (UTC)
Clearly, the decline of Low German speakers in Brandenburg and Saxony-Anhalt could be observed in the last decades. But this process also happened and still happens (to a more or less extend) in other regions of Northern Germany, where traditionally Low German was spoken. That the original Low German in Saxony-Anhalt had been replaced by Upper Saxon Dialects is quite a remarkable claim with no scientific proof. So far I could not find any source, which would document that. The reference mentioned above does not even mention the Benrath Line at all - nor does it try to draw a current location of this line. Additionally, it does not make any remarks on the situation in Saxony-Anhalt. The sentence in this article "In the second half of the 20th century, a northward movement of the eastern end of the Benrath line has been observed" is merely a popular assumption due to the fact that there is a growing influence of the Berlin colloquial language on the Brandenburg surrounding.--Zarbi1 19:12, 6 November 2007 (UTC)

OK, I've removed that. Clearly the rationale for it has been so far undermined that it has become an entirely doubtful statement. The increased use of standard language in an area does not equate to a shift of dialect boundaries. If anyone thinks this should be kept, they need to find some sources. --Doric Loon 13:05, 8 November 2007 (UTC)

Maybe, it would be interesting to point out, that to a large extent the data and the fixation of the line dates back 100 years (which gives room for speculations that the map shown is outdated) Well, unfortunatly it seems to be that there is no recent material.--Zarbi1 14:00, 8 November 2007 (UTC)
That's not just interesting, but the clue of the whole story. The entire concepts doesn't only rely on data that is over 100 years old but also on the academic interpretations of scholars from that era. Since than it has never really been reassessed. --El bes 14:46, 8 November 2007 (UTC)
It's a shift in dialect boundaries. In parts of East Germany Low German has been replaced by Berlin dialect (or varieties close to it) which are classified as Middle German, the process has been going on for about 50 years. Increased use of standard language may also be going on at the same time, but that is not the rationale for stating a movement of the boundary. It's a fairly recent phenomenon for which there are few scientific source, but a couple do exist. I added this link earlier which is a good source, I don't know why it was removed. Anorak2 (talk) 11:59, 5 February 2008 (UTC)

Symbols in Phase 1[edit]

Angr, the symbol you have put in for the ae-ligature with a lengethening sign appears on my screen as a box. We had this before on another page and you found a way to fix it. Any chance you can do your magic here too? --Doric Loon 20:46, 11 Jun 2005 (UTC)

I added the {{Unicode}} template around them. Does it work right for you now? (If not, why not switch to a real browser like Netscape, Mozilla, or Opera?) --Angr/tɔk tə mi 21:44, 11 Jun 2005 (UTC)

Yup, that's fixed it. Good stuff! --Doric Loon 22:56, 11 Jun 2005 (UTC)

dates of the shift[edit]

i wonder how general the agreement on the given dates is. the text as written implies that there must necessarily have been a long period between each of the stages, as otherwise sounds would have merged. but this is not necessarily true; multi-level chain shifts can happen over a very short period of time (e.g. the great vowel shift is four levels, < 150 years), and there is no need for one shift to "stop" before the other "starts". there are various modern cases where multi-level shifts occur simultaneously and move from dialect to dialect as a unit. the late date of the fourth shift appears well documented, but it's possible the other three occurred nearly simultaneously but took a long time to spread; this is what is implied in Waterman "A History of the German Language". Benwing 05:55, 28 July 2005 (UTC)

I'm not sure that Waterman is the best source, but of course it is possible that phases 2, 3 and 4 could have occurred fairly rapidly. The text as written does not imply anything else. It does say we have to allow for the possibility that some phases MAY have lasted longer, but the dates given in the table place phases 2, 3 and 4 at least partly in the 8th century. I think the text as written also gives due attention to the fact that the dates are tentative. If you want to write more on the mechanisms in general, that might be good, but the place to do it is in the article chain shift.--Doric Loon 09:56, 28 July 2005 (UTC)
Waterman actually believes that phase 1,2,3 were close together and phase 4 much later and perhaps should not really be considered part of the shift at all; there is some logic to this, since phase 1,2,3 affect successively smaller regions whereas phase 4 affects everywhere, including places not affected by phase 1, and affects only a single series, whereas the other affect (to some degree) all three of p/t/k. Benwing 01:32, 29 July 2005 (UTC)
Well of course the article already notes that not everyone groups phase 4 under this heading, for precisely these reasons. But it is better covered here than in an article of its own. I think it goes beyond the scope of this article to list all the details of every dating theory, so I would not be inclined to add much at the moment. Better to give one plausible set of dates and be content with the warning that these are tentative. But a sentence at the bottom of the chronology section to say that "an alternative view will be found in Waterman" might be quite useful. --Doric Loon 09:31, 29 July 2005 (UTC)
OK, I went ahead and did that. Perhaps you could add a page number, and put full details of Waterman at the bottom of the page?--Doric Loon 09:44, 29 July 2005 (UTC)


DAB, do you have more information on that Lombardic reference? For example, which of the four phases you are talking about? Or a reference? I suspect that doesn't belong under chronology, but you can tell us more, perhaps. --Doric Loon 13:00, 10 August 2005 (UTC)

Notation of the overview table[edit]

We presently have the sound shifts described in the Latin alphabet, which is fine, except that if you're not already familiar with German orthography the "ch" in kch might be confused with other sounds, and even if you are familiar with German, the hh is not particularly clear. I'm tempted to replace them with IPA, possibly with links to the relevant sounds (something like this: þ), but not speaking German of any variety I'm reluctant for fear of stuffing it up. Moreover, IPA might not be the best option here, but things do need to be better explained. What do people think? J.K. 12:57, 22 October 2005 (UTC)

(I've been just a reader of the article.) The usage of some traditional Latin notations looks fine to me, but I would put some short description on the correspondence of the used notation to sounds in front of the table where it is used. The description could consist of notes like this one:
In these OHG words, <z> stands for a voiceless fricative that is distinct somehow from <s>. The exact nature of the distinction is unknown; possibly <s> was apical while <z> was laminal.
from High German consonant shift#Phase 1 (so, I'd move it up). (And here it is obvious that IPA is not a suitable alternative here, since the exact features of the sounds are not known.--Imz 19:41, 14 November 2005 (UTC)

Maybe I'm alone here, but wouldn't it be helpful to have OHG and modern cognates in the chart? The formula *t→ss, for example, is problematic for me, as my favorite OHG dictionary (R. Schützeichel, 1995) has zz (z.B. ezzen), a further, later change being required for essen. Or maybe it's the column heading High German Shift Germanic→OHG that is misleading me... Varoon Arya 00:18, 7 November 2007 (UTC)

GA nomination failed[edit]

It failed on account of poor referencing or no inline citations. Lincher 23:59, 4 June 2006 (UTC)

what do we need {{GA}} for if by now GAs essentially require {{FA}} qualities? Also, no review or nomination is linked, you simply removed the tag. I haven't contributed to this article, and I think it easily qualifies as "good". Hence, I will re-add the template, until some formal "review" or whatever is required in GA bureaucracy these days is undertaken. dab () 20:06, 21 July 2006 (UTC)

Tag bad example[edit]

I'm not a linguist, but maybe the example of the third shift is not the best one, since Dutch also shifts for this example. (but probably not on the whole)

           11:15, 8 June 2006 (UTC)
Don't unerstand, all the examples of phase three of the shift show Dutch unaffected. Phase four, on the other hand, does affect Dutch. The examples seem fine to me, but do come back on that and say which example is bothering your. --Doric Loon 12:07, 8 June 2006 (UTC)


comments are welcome at Talk:History of German: should the article's scope extend to Low German, or is the history of German the history of the 2nd sound shift? Also, how should the ToC ideally be arranged? High and Low in separate h2 sections, or Early and High Middle Ages in separate h2 sections? dab () 20:10, 21 July 2006 (UTC)

Would a discussion of the relationship to the Ingvaenoic shift make these matters clearer?

I think not. By all means add links. But this is a detailed article on one particular phenomenon. It has too much detail for a general history of the German language article, and will only be cluttered by having bits of other things merged. I think we need a good general overview article on historical German philology, linking to deep but narrow articles on particularly influential or complex events. This article is the latter. --Doric Loon 12:27, 7 December 2006 (UTC)

/θ/ > /d/[edit]

This is not part of the HG consonant shift, because:

  1. it is not restricted to High German dialects
  2. it is phonogically the opposite of the sound shift, since it is a lenition
  3. it does not form part of a series with the other voiceless fricatives, as the voiced and voiceless stops do.
  4. it does not show the S->N diminution seen in the effects of the sound shift
  5. most handbooks do not treat it as "part" of the Sound Shift, even though they accept a "connection" in the HG dialects. Waterman's view is not unrepresentative at all, as far as I can see.

I quite accept the case for covering /θ/ > /d/ on this page, since it modifies the effect of the sound shift on the HG consonant system, and it hardly merits its own separate page. But I would say that presenting it as "phase" of the sound shift does not represent the communis opinio. --Pfold 18:41, 19 December 2006 (UTC)

Agreed. It's interesting and it occured in the same timeframe, but it is very different. It doesn't have a stop becoming a fricative/affricate or devoicing, in fact, it has the opposite in both terms.Cameron Nedland 22:43, 27 June 2007 (UTC)

Not sure about that. Some text books take them together, others don't. The article notes the uncertainty. --Doric Loon 12:55, 28 June 2007 (UTC)

"Taking them together" is perfectly reasonable - they both fall under the heading OHG consonant changes. Treating /θ/ > /d/ as a stage in the sound shift, on the other hand, suggests a level of phonological ignorance that one would not expect in a textbook. --Pfold 13:50, 28 June 2007 (UTC)

Ignorance is a strong word for an alternative view. This is a chain shift. To see /θ/ > /d/ as the next stage in the chain after /d/ > /t/ is not illogical. It depends how narrowly you define HGCS. One the one hand, the three "tripple shifts" are very neat by themselves, and there is something to be said for taking them together and everything else separately. On the other hand, since the HGCS is not in any case ONE shift, but rather a series taking place over 5 centuries, there is no real reason why it should not mean ALL the consonant changes in High German in those centuries. And since the geographical distribution of them varies, the fact that this one goes further north is no reason to exclude it: to argue from the modern languages (status of Low German) would be anachronistic. I might mention that a linguist I spoke to recently (who wrote his Habil on Germanic etymology) referred to the the /v/ > /b/ shift in haben and geben as HGCS and explained that he saw no point in treating it separately from the other consonant changes of the period. I've no axe to grind here, and if you want to change the prioritisation in the article, that's fine. (I mean, at present the article takes "phase 4" as the norm and says there are other ways of looking at it - you are welcome to reverse that, though it may be more work than you think!) But don't characterise the other view as wrong. --Doric Loon 09:02, 29 June 2007 (UTC)

I think the changes you've made to the article have dealt with the issue very well. --Pfold 10:27, 29 June 2007 (UTC)


I have just removed this mnemonic from the article:

Ich habe einen Rechen, ein Schiff, einen Bruder und einen Apfel, und ich trinke Salzwasser.
(English: I have a rake, a ship, a brother and an apple, and I drink saltwater.)

This was partly because it was out of place (between the table and its footnotes) and partly because it seems rather trivial and clumsy. But if people really are using it as a teaching aid (if there is a source for it?) then it perhaps should indeed go in somewhere. Presumably the ch of ich should also be bold. And the b of habe, which User:Cameron Nedland de-bolded should be bold if the /v/ > /b/ shift is regarded as "in the same context" - which is presumably what the originator of the mnemonic intended. --Doric Loon 15:15, 29 June 2007 (UTC)

True, but nothing on the page mentioned anything about a v→b change.Cameron Nedland 07:26, 30 June 2007 (UTC)

/v/ > /b/[edit]

Since the shift /v/ > /b/ has now been mentioned a couple of times here, should it go in the article as a related change? Calling it HGCS is unusual (my friend mentioned above is untypical, though certainly not uninformed) but since we are unlikely to find any other article in which this shift can more meaningfully be discussed, we could add a brief section on other consonant changes of the period. As far as I can see, this was a fairly limited shift, which applies only in medial position. In German it applies only intervocallically, but in Dutch it seems to apply also when there is a following j. So:

West Germanic German Dutch English
*gevan geben geven give
*hevjan haben hebben have

(This is spelling the reconstructed WG words with v, which is not usual, but I can't find how to enter the b with a stroke through it; at any rate, the philological value is v!)

Another consonant change in High German at this time was the hardening of g to a stop, where in WG (and in Old Saxon and modern Dutch) it was fricative. (In English it was palatalised to y.) All this will need some more thought before it goes in the article, but what do you think? There certainly is no-where else at present where we can write it up. --Doric Loon 15:34, 29 June 2007 (UTC)

Perhaps this should be included in step 4, after all, this is another case of a fricative turning into a stop.Cameron Nedland 07:29, 30 June 2007 (UTC)
It belongs with Phase 3: In WGmc each of the voiced stops had a voiced fricative allophone in certain environments. In the case of the dental, the fricative became a voiced stop and then mostly became devoiced with the rest of the [d]s. In the case of the labial, the fricative became a stop in most dialects, but in UG frequently went on to be devoiced (depending on environment), so there are forms 'leven', 'leben' and 'lepen' attested. Similarly for the palatal. As I remember it, though, the details are rather messy, not least because of the problems of the MS spellings.
So strictly speaking it's not /v/ > /b/, it's /b~v/ > /p/ or /b/ or /p/&/b/ or /b/&/v/, according to dialect. --Pfold 09:08, 1 July 2007 (UTC)

OK, I've added an "other changes" section, which you will probably want to expand, though I don't think these want long and detailed discussions; they are peripheral in the article. --Doric Loon 21:25, 1 July 2007 (UTC)

The statement that 'Farbe' shows a /v/ > /b/ change is only correct if a /w/ > /v/ change had already taken place. Given the tiny number of examples (I could only think of gerben), I would be surprised if there's evidence to support this. It needs to be sourced or the statement dropped. To be honest, I don't see the point of including this info anyway, as it has nothing to do with the sound shift. --Pfold 13:23, 22 October 2007 (UTC)

Many respectable texts give Early PGmc [β] as an alternate sign for the [v] or [b~v] reflex of PIE [bʰ] discussed above. In the absence of the 'stroke-b', shouldn't we be using this instead of the less precise [v]? Varoon Arya 00:58, 7 November 2007 (UTC)

If you like, but I'd prefer to get the stroke-b symbol. Can anyone help? --Doric Loon 09:39, 7 November 2007 (UTC)

I found this in the article West Germanic languages: ƀ. I'm not sure how it got there, as I still don't see it in the list. But it could be copied and pasted from here, I suppose... Varoon Arya 17:18, 7 November 2007 (UTC)

Thanks. I've put that in. --Doric Loon 13:03, 8 November 2007 (UTC)

More or Less Shift in Standard German[edit]

Standard German has some words which have been shifted further than listed in the table (trock- in trocken, cf. Dutch droog, Old English drȳge, English dry), and some less (bitter, finden, binden, cf. Dutch bitter, vinden, binden, English bitter, (to) find, bind). This probably is due to SG being the result of assimilation and compromise of various dialects with varying degrees of the HGCS. It seems in some sense misleading to say that standard German has or doesn't have each shift. One should perhaps add qualifiers like "usually" "rarely", "generally not", etc. -User: Nightvid

I am not sure about that. I suspect that the first three phases admit no random exceptions. (The fourth is messy, as our ongoing discussions show.) Finden and binden may well be systematic exceptions - a fundamentally different thing - since proximity to a nasal often blocks sound shifts. Bitter might be a Low German form; the article already has a section on this phenomenon. I am on holiday now, without my books, so I can't check the etymologies of your examples, but I don't think peppering the article with weasle words would help here. --Doric Loon 12:43, 30 July 2007 (UTC)
Kluge says the unshifted /t/ in bitter is because of the following /r/. Fintan and bintan are certainly found in OHG - shifts of /bdg/ to /ptk/ were partly reversed later on because of Frankish influence on UG, according to Wells. HG is phonologically based on CG rather than UG, so this is what you'd expect anyway. The fact that some words in HG have non-CG features is matter of the lexicon, not the phonology. --Pfold 21:19, 31 July 2007 (UTC)
Standard German is not really based on Central German. If it were, we would have to say "Perd" and "Kopp" instead of "Pferd" and "Kopf". Standard German is a compromise between Central and Upper German. The ostensible exceptions "finden" and "bitter" have already been clarified. As to trocken: This is not quite the same word as Dutch droog. The equivalent form is Middle High German trüge (which still exists for example in my Rhineland dialect as "drüch" = trocken). Trocken is a derived form of trüge in which the -g- was doubled and therefore regularly changed to -ck- (like Dutch "brug", German "Brücke"; or Dutch "rug", German "Rücken"). — Preceding unsigned comment added by (talk) 01:29, 7 January 2013 (UTC)

Phase 4 again[edit]

I am a little worried about Cameron Nedland's changes. I must admit I toyed with this idea myself a couple of weeks ago, but decided against it. Of course, it makes the whole thing much neater, but I suspect it is cheating. For one thing, þ is a VOICELESS fricative. It does have a voiced allophone ð in Germanic (used medially), and both shift to d. But the shift v to b does not affect the allophone f (i.e. it only occurs medially). And Germanic "gh" (sorry, I don't thave the right symbols) did not have a corresponding allophonic relationship to "ch". So this whole thing is not nearly as tidy as you would like it to be. I would be happier if you could find a reference for treating these three things together. And you certainly need to build a fuller discussion into the phase 4 part of the "in detail" section. --Doric Loon 12:43, 30 July 2007 (UTC)

I'm not worried at all: any claim that the voiced allophone is becoming a voiced stop in the 9/10 C is refuted by 8th C Bavarian forms like 'kepan' for "geben", which indicate that the stop and fricative allophones must have coalesced before or during stage 3 in at least one dialect. And of course we know this must have happened to the voiced dental fricative, since it became /t/ everywhere in stage 3. --Pfold 23:09, 30 July 2007 (UTC)
Well, I'm sorry, but otherwise the stage 4 just makes almost no sense. Could thorn have been voiced at this time (through initial fricative voicing)? And I don't see why a corresponding relation with the voiceless fricative in the /ɣ→g/ change would have to do with this. Please tell me if I am wrong.Cameron Nedland 17:48, 31 July 2007 (UTC)
There are several issues here.
  1. It's not enough for what you've put to be a sensible suggestion - it has to come from the published literature. Otherwise it counts as original research.
  2. The fricative [ɣ] is voiced not voiceless. Why would one expect both voiced and voiceless fricatives to changed to voiced stops at the same period?
  3. It seems very clearly invalidated by the evidence of early texts. If this account is to stand, it cannot do so without explaining why these shifted forms are found in the texts.
  4. Looking at some of the handbooks, I see that the formulation here is not right anyway. Those I've checked (von Kienle, Wells, and Penzl) take the voiced fricatives as the Gmc phonemes, with /b d g/ developing as allophones in WGmc. In most HG dialects, this allophonic variation then disappears (i.e. all the voiced fricatives become stops) and clearly it must do so before the shift of the voiced stops to voiceless stops, most clearly seen in /ð~d > t/.
  5. This means the development of the fricatives must be regarded as part of Stage 3, not a later development.
More generally, I dispute that many Germanists would regard /þ > d/ as part of the sound shift, i.e. there is no Stage 4, so the fact that it makes no sense is not a problem! If /þ > d/ was part of the sound shift, surely we should expect the voiceless fricative /f/ to undergo a parallel development to /b/ - this just shows that /þ > d/ is the opposite of the sound shift in phonological terms (never mind that its geographical distribution bears no relation to the sound shift either). --Pfold 20:21, 31 July 2007 (UTC)
Pfold, this is partly quite useful.
I think your objection to the idea of "phase 4" is beside the point: the first three phases do not all go in the same direction, therefore four cannot be opposite: the first three phases all have different geographical distributions, therefore it is irrelevant that 4 is different again. We have already established that the term HGCS is used in two different ways: some people mean just the very neat set of 3x3 shifts, which is rather pretty but means you shut your eyes to everything else going on at the time; others take it to mean the consonant changes which mark High German off from Germanic, which is rather more useful and comprehensive, but less neat. In true wiki style we have reached the compromise of talking about the core group and the related changes which belong in the same context, and with that caveat, I think there can be no difficulty at all about the terminology of phase 4.
However, I have two problems with the way Cameron Nedland uses the term phase 4. The first is chronology: I am not convinced all of these changes happened after phase 3, which invalidates the terminology. /þ > d/ did take place in the 9th century: that is not in dispute, because unshifted forms are found in 8th century texts. This applies also to the allophone ð, as we see in early OHG bruother → classical OHG bruoder; and here, Pfold, you are wrong to speak of /ð~d > t/; phase 3 /d > t/ precedes phase 4 /þ~ð > d/, therefore we never experience /ð > t/. However, the same is not true of /v > b/, as Pfold correctly shows: Bavarian kepan for "give" shows /v > b > p/, indicating that /v > b/ predates phase 3.
My second problem as I said earlier is that the phonetic conditions under which we see /þ > d/ are quite different from those for /v > b/. And that speaks against the grouping of three which Cameron Nedland is attempting.
In short, what I think BOTH of you are guilty of is trying too hard to make a complex situation look neater than it is. You need to go back to the text books and get dates and data on the g and b shifts, and then discuss them properly from the literature and not from your own hunches. --Doric Loon 21:48, 31 July 2007 (UTC)
Dammit. You're right.Cameron Nedland 01:05, 3 August 2007 (UTC)

On the other hand, this discussion has raised a couple of points which do need to be explained, so thank you both for sticking your oars in. I will attempt a fix, not a revert. Tell me if you think it can be done better. --Doric Loon 18:50, 3 August 2007 (UTC)

OK, that's all I have time for, as I'm going off camping for a week. Ergo, no internet (yippee!). Some of this needs fuller explanations and examples. Please give it a go. --Doric Loon 19:44, 3 August 2007 (UTC)


As for the ɣ→g, it says "Dutch has retained the original Germanic /ɣ/, though as Dutch spells this with <g>, the difference is invisible in the written form." It's not clear however what 'the difference' is. Difference between a ɣ and another sound spelled <g>? Which one? Dutch has, I believe g→ɣ, so exactly the opposite. So maybe it's the difference between the orginal ɣ and the ɣ from /g/? If so, I think the whole sentence can be removed. Jalwikip 14:21, 6 August 2007 (UTC)

The difference referred to is the difference between English/German and Dutch, whereby Englihs and German both have g (historically an innovation) whilst Dutch has (the original Germanic) ɣ. When you look at English good, German gut, and Dutch goed, you can't see that there is a difference, and people tend to take philological changes more seriously when they can see them in the spelling. Thus we often simply hear that Dutch pronounces g as a fricative, whereas we no-one would say that German pronounces t as tz. It's as though we only half register a phonological shift when the spelling hides it. Which is why I drew attention to it that way. But feel free to rephrase. --Doric Loon 14:11, 8 August 2007 (UTC)
But just to correct you in point of fact, Dutch never experienced g→ɣ; the Germanic ɣ came down into Dutch unchanged. Dutch never had a g at all. The Germanic word for 'good', in so far as we can tell the precise phonetics of a reconstructed language, was pronounced identically to Dutch goed. --Doric Loon 14:15, 8 August 2007 (UTC)
Ok, just my error then. Jalwikip 13:05, 23 August 2007 (UTC)
Just a late correction to your statement Doric. It's only modern Dutch that lacks [g]. Middle Dutch inherited it from Proto-Germanic in its original allophonic distribution: g was a fricative normally, but a plosive when geminated (e.g. secgen "to say" which shows this) and in the combination /ng/ (e.g. coninc "king" where it was devoiced word-finally). I believe many Low German dialects retain this final-devoiced -ng as well. CodeCat (talk) 19:54, 11 April 2012 (UTC)


Should it be noted that Yiddish went through the same shift and the only major difference in the effect now is that Yiddish has merged /pf/ with /f/?Cameron Nedland 20:26, 9 September 2007 (UTC)

Why not? But only in passing, let's not have Yiddish versions of every cited form (the Low German ones are already cluttering the article rather than enriching it). Some Central German dialects also merge /pf/ with /f/ - I have a friend from Lüneburg who says Ferd for Pferd. --Doric Loon 21:43, 9 September 2007 (UTC)
Well, should we have a note saying something like "Some central German dialects and Yiddish went through the /p/→/pf/ change, but have since merged /pf/ with /f/ word-initially"?Cameron Nedland 03:32, 10 September 2007 (UTC)


The article states that "Kch" is used instead "K" of in "Southernmost Austro-Bavarian and High Alemannic," with the example "Bavarian: Kchind." I think this is incorrect. Kch is used ony in Alemannic, not in Bavarian. In Austria, it used in Vorarlberg, where Alemannic is spoken, and not in the rest of the country, where Bavarian is spoken. This includes the southernmost areas of Styria and Carynthia. 12:42, 24 September 2007 (UTC)

Do you have a source for that? I am a little afraid that in the dialect question people are putting in what they think they know from granny. Please only put into this article what you find in a reputable book on the subject. If you have a source, feel free to correct. --Doric Loon 17:13, 24 September 2007 (UTC)
I checked the sources, and things are not quite as I said. The Kch is used in Tirol (Kchadl = Kerl, Kchoich=Kalk)and Carynthia (schluckchn as well as schluckhn = schlucken but Muckn (Muecke) and Pruckn (Bruecke)). In Styria the letter k is merely aspirated (Khloan=klein). Thus, the Kch is not as pronounced as in Alemannic, but still exists. (Hornung, Roitinger, Die Oesterreichischen Mundarten, Eine Einfuehrung, Oebv&hpt, 2000) Interestingly, Merkle's Bairische grammatik (DTV, 1976) does not mention this at all. I hereby withdraw my comment. 19:59, 24 September 2007 (UTC)
First of all, sorry for my poor English. You can find some maps in the "dtv-Atlas Deutsche Sprache", p. 64. The "Mittelhochdeutsche Grammatik" by Paul et al. provides some more information, e. g. § 88 (general overview about that shift), §159,1 focuses on Bavarian, § 160 on Alemannic. In MHG even Middle Bavarian was affected. Today there are two pronounciations [χ] and [kχ], I will just quote the Mittelhochdeutsche Grammatik (24th ed.), p. 167: "Heute wird nur im Südbair. und Hochalem. [kχ] (bzw. daraus hervorgegangenes [χ]) gesprochen, während nördlich behauchter Starklaut [k'] anschließt. In mhd. Zeit hat [kχ] sicher weiter nördlich gereicht als heute, wahrscheinlich noch das ganze Mittelbair. umfaßt." Hope that helps. -- (talk) 14:05, 23 November 2007 (UTC)
Wait, kχ or kx?Cameron Nedland (talk) 22:05, 23 December 2007 (UTC)

The whole concept is weak[edit]

My question is: Which primary sources are the base for the whole consonant shift concept and the quoted stages and geografic distribution? Or was it postulated ex post?

My information is, that the very oldest text in any form of Old German is the Abrogans dictionary, that translates several hundred latin words into either Old-Bavarian (missing original) or Old-Alemannic (3 copies are conserved). As a matter of fact, the Abrogans manuscript already shows the post-consonant-shift situation, eg: friuntscaffi (friendship).

  • How is it possible, that in a change from A --> B, we find the form B in the very oldest text?
  • Is Abrogans the oldest scientifically exploitable Old German text?


If Old-Bavarian together with Old-Alemannic and Langobardian, are the origin of the consonant shift, and Old-Saxon and Old-Frankonian did not share this language characteristics, why are they put together in one group that is subsumed as "Old-German" and the southern germanic languages are labelled dialects?

This categorization was done by romantic-nationalist German linguists of the 19th century, like the notorius Grimm brothers among others, and was taken over by English language linguists without rechecking the data.

Today English-language linguists tend to peel out Old-Saxon from this unluckily choosen "Old-German" super-group and put it alongside Old-Frisian and Old-English.

Dutch linguists tend to peel out lower Frankonian from the "Old-German" super-group, labelling upper Frankonian together with Bavarian and Alemannic as Old High German.

But what is true and what is unbiased scientifically neutral and non-nationalistic information? Shouldn't the whole 19th century terms and classifications all together be reevaluated?

--El bes 01:52, 23 October 2007 (UTC)

Well, certainly not here. This is not the place for new research. There is no doubt that there is a dialect continuum from Graz to Flensburg, and that that was far more obvious in the 8th century than it is today. Political discussions about what to call them are rather boring. This article is about the way sound changes rippled through these dialects, and that is interesting, and to make it discussable we'll just use the most familiar names for dialects. --Doric Loon 04:38, 23 October 2007 (UTC)
quod erat expectando ... and P.S.: by discussing you are doing new research! --El bes 04:44, 23 October 2007 (UTC)
I think you need to read Wikipedia's original research policy to understand what that means. --Doric Loon 12:42, 23 October 2007 (UTC)
Doing personal research in Wikipedia is not allowed, I know. But relying on outdated and questionable research form 19th century scholars IS allowed. Is that what you want to say? --El bes 12:49, 23 October 2007 (UTC)
Of course not. If you have published material you want to show us, we'll look at it. But if it is a new approach to classifying Germanic langauges / dialects, this is not the article to put it in. HERE I would want to concentrate on existing literature on this particular shift. If linguists working on this shift have applied your preferred categories to it, then we are in business. Possibly. --Doric Loon 20:34, 23 October 2007 (UTC)
I will try to find some new English language scientific sources on that. --El bes 00:07, 24 October 2007 (UTC)
German language sources are also fine for this article. --Doric Loon 13:27, 24 October 2007 (UTC)
Before I start searching, is Abrogans the oldest text? Is Abrogans the basis for the chronological fixation of the consonant shift or not? What is the basis? Some minor glossae before Abrogans? Which glossae? Which germanic texts, beside the gothic ones, are dated 750 or earlier? This is not a rhetorical question, I am really interested. --El bes 20:39, 24 October 2007 (UTC)


Some scholars have compared the sound shifts in German with SOMETHING vaguely similar in Celtic... I cannot recall the details, and cannot yet find the link. Does anybody know what I am vaguely remembering? Around the time of the Grimm's Law shift / HGConsonant Shift, there was some marked and vaguely similar shift in Celtic, which some scholars have linked with those (more pronounced) changes in Germanic. —Preceding unsigned comment added by (talk) 22:22, 16 October 2007 (UTC)

Celtic did undergo a sound change in which, among other things, 'p' did change to /φ/, which is vaguely similar to /f/.Cameron Nedland 18:28, 23 October 2007 (UTC)

Question about t > ts[edit]

I don't understand why that shift only affects Upper German. You can find this change also in Central German dialects, e.g. Zeit (zidd, zit, etc.) (engl. time), sitzen/setzen (engl. sitt). It seems to affect almost all CG dialects according to I have no idea if it does affect all words but certainly this shift also exists in CG dialects. Is there any reason why only UG is mentioned here? -- (talk) 15:00, 23 November 2007 (UTC)

This was taken from the standard textbooks. If we have misread them, or if sources disagree, please find a textbook which says what you think, and then we can adjust the text accordingly. I would be very reluctant to change it just because of your personal observations, though. --Doric Loon (talk) 08:53, 24 November 2007 (UTC)
Actually the current formulation about concentration in UG is thoroughly misleading. The fact that it's only in UG that all three of these changes take place is correct, the implication that /p > pf/ and /t > ts/ are themselves somehow concentrated in UG is utterly wrong. --Pfold (talk) 10:04, 24 November 2007 (UTC)
Well, that's just what I meant. I've just provided a source above, i. e. the Digitaler Wenker Atlas, which is somewhat reliable (cf. Niebaum, Hermann; Macha, Jürgen: Einführung in die Dialektologie des Deutschen, 2nd ed., Tübingen 2006, p. 68 - 70.) The Mittelhochdeutsche Grammatik (24th ed.) (§88, p. 117f.) just mentions five unshifted words (Central Franconian) dat, dit, it, wat, allet. However that's just the coda. The onset is shifted: "Germ. stl. Verschlußlaute /p, t, k/ werden als Geminaten oder im Anlaut oder inlautend nach Konsonanten /l, r, m, n/ zu den Affrikaten /pf/, /tz, z/ [ts], /kch/ verschoben. Nur /t/ > /tz, z/ gilt im ganzen hd. Gebiet: mndd. [i. e. Middle Low German]: tal, swart, sitten - mhd. zal, swarz, sitzen". Another source (referring to Old High German) would be Schmidt, Wilhem: Geschichte der Deutschen Sprache, 9th ed., Suttgart 2004, p. 204 f.: "Die Verschiebung von /t/ > /ts/ ist gleichmäßig über das hd. Gebiet verbreitet. Dialektunterschiede gibt es hier nicht. [...] Im Mittelfränk. wurde das auslautende /t/ der vier neutralen Pronomen that, it, wat, allet nicht von der Verschiebung betroffen. Mittelfränk. heißt es demnach: that waʒʒar." -- (talk) 22:34, 24 November 2007 (UTC)
OK, it's possible the sources we used referred to the phases as wholes and didn't distinguish the parts of them. If you have more precise information, please do make the corrections, and add citations in footnotes. Thanks. --Doric Loon (talk) 11:59, 25 November 2007 (UTC)
Actually, there is really no need for citations here - the correct details, which are here misrepresented, are in every book on the subject. The only mystery is why we didn't all spot this error before! --Pfold (talk) 17:30, 26 November 2007 (UTC)
Well, actually my English is not that good. Maybe you could do that? The precise information would be something like that: In CF only "wat, dat, dit, it, allet" are unshifted. That could be mentioned in a footnote. The citations would be: Paul/Wiehl/Grosse, Mittelhochdeutsche Grammatik, 24th ed, Tübingen 1998, 117f. (§88). Niebaum/Macha refer to today's spoken dialects: "Das Ripuarische und das Moselfränkische zeigen gegenüber dem Rheinfränkischen (und Hessischen) unverschobenes t in den Prominalformen dat, wat, it, dit ('das, was, es, dies'), z. T. auch in allet ('alles') und in der Adjektivendung des Neutrums (z. B. schönet 'schönes')." (Niebaum/Macha [quotation see above], 222). That would be more precise. However, in my opinion the Mittelhochdeutsche Grammatik is sufficient. -- (talk) 13:58, 25 November 2007 (UTC)
The more exemptions you find, the more you prove how weak the standard model is (or the standard classification of what is labelled a "high" or "low" variant and the word "High German" itself. Thanks for mentioning that. --El bes (talk) 17:03, 25 November 2007 (UTC)
Sorry, but these are examples of t > ss not taking place & nothing to do with t > ts --Pfold (talk) 17:30, 26 November 2007 (UTC)
OK, I have now revised the Phase 2 section. Once I came to look at it I found a number of other errors, which I have attempted to fix. I have replaced the arrow → with the more traditional >, and will do likewise in the other sections, when I have another spare moment.
One other issue I'd like to raise is whole question of using "phases" to present this material. All the lit. I've looked at treats our first two phases as a single event with two aspects. The main reason is that what is really happening here is that each stop is developing three allophones which end up as: the unaspirated stop (after fricatives), the affricate, and the geminated fricative. Since the latter are simply two developments of an earlier aspirated form (still present in English), presenting these as separate "phases" is misleading. IMO we should present the raw data of these changes, then present material about their nature and chronology - which are interpretations of the data. Casting the changes of the stops in 2 "phases" may seem helpful but is not what literature says - and it is arguably not NPOV since it says these are two historically discrete events. And of course, to return to "phase 4", calling it that prejudges the issue of whether it really is part of the sound shift.
I propose, therefore, that we should offer a purely descriptive account of the changes to start with(i.e what happened to the Gmc voicelss stops, voiced stops, and /th/ - taking the starting point rather than the ending point is more NPOV, since there is less interpretation), followed by sections on chronology, phonological interpretation, geographical variation etc. etc. --Pfold (talk) 18:30, 28 November 2007 (UTC)
Sorry, but these are examples of t > ss not taking place Yes, but it's the only phase, where you can find unshifted t in CG: dat, wat etc. That's why I mentioned that. All other ts are shifted (ts or ss).-- (talk) 20:16, 28 November 2007 (UTC)
Yes, I see! --Pfold (talk) 21:19, 28 November 2007 (UTC)

Census of the literature[edit]

Here's a quick census of how the sound shift is presented in the handbooks. Obviously this is a very crude summary, but it confirms my earlier gut feeling that most scholars present the SS as having two main parts (voiceless and voiced) and that th>d is not regard as a "phase" of the SS.

How many "phases"?

One section for the voiceless and one for the voiced stops:-

  • Bach
  • Braune/Reiffenstein
  • Eggers
  • Gerh. Wolff
  • Keller
  • Moser/Wellmann/Wolf
  • Penzl (1971 & 1975)
  • Russ
  • Sonderegger (1979)
  • v. Kienle
  • Voyles (1992)
  • Wright (1907)

Note: Eggers & Voyles restrict the term Sound Shift to the voicelss stops

Three sections (our phases 1-3):-

  • Chambers & Wilkie
  • dtv-Atlas
  • Waterman
  • Wells

th > d

not mentioned at all:-

  • Bach
  • Eggers
  • Gerh. Wolff
  • Sonderegger (1979)
  • Waterman

treated in a separate part of the book (not under the heading "sound shift", often dozens of pages away, sometimes with other fricative changes):-

  • Braune/reiffenstein
  • Chambers & Wilkie
  • v. Kienle
  • Wright (1907)
  • Voyles (1992)

treated as "linked" with the SS but not part of it:-

  • dtv-Atlas
  • Keller
  • Moser/Wellmann/Wolf
  • Wells

covered with voiced stops:-

  • Penzl (1975)

covered in a separate chapter immediately before or after the voiced stops:-

  • Penzl (1971)
  • Russ

(but neither refer to it as part of the SS)

Please add more! --Pfold (talk) 21:19, 28 November 2007 (UTC)

Braune/Eggers (1987) discusses the changes from WGmc þ in several sections: Hochdeutsche Lautverschiebung (§85), Althochdeutsche Konsonantenschwächungen (§102), Notkers Anlautgesetz (§103) and in great detail under Geräuschlaute (§165-168). Meineke/Schwerdt (2001) discusses the change specifically under the title Spirantenschwächung; it follows immediately after the discussion of the 2nd SS and in the same chapter. I don't know where you would fit that into your list, but I hope the info helps.

In general, I would advise against putting too much into periodization regarding the 2nd SS. I'm thinking specifically of Schwert (2000) who argues (rather well, IMO) that the 2nd SS itself is a faulty concept, and that instead we should be discussing consonantal changes of individual WGmc dialects, in particular Early Old Bavarian, Early Old Alemmanic and Early Old Frankish. In other words, rather than a true SS like the one described by Grimm's Law, we have a group of tangent yet isolatable areal shifts. I realize that it is too early to scrap the 2nd SS altogether; besides, it is a useful - not to mention deeply entrenched - concept. But Schwert's evaluation is is spot on as far as trends in modern Germanic studies go (cf. van Coetsem, Voyles, Davis, etc). Aryaman (☼) 23:10, 28 November 2007 (UTC)

Do you mean Schwerdt (2000)? Has anyone looked at Schwerdt, Judith, Kontroverse um die 2. Lautverschiebung? --Pfold (talk) 11:19, 29 November 2007 (UTC)
Yes, Schwerdt. (Sorry about the typo. The first reference [Meineke/Schwerdt (2001)] was the correct one and is the same Judith Schwerdt.) I haven't yet seen the book you mention, but I have read a review and long excerpts from her Die 2. Lautverschiebung: Wege zu ihrer Erforschung (Heidelberg 2000), the conclusion of which I was referring to earlier. (In connection with the general topic, Iverson/Salmons (2006) from the Journal of Germanic Linguistics [1] presents an interesting treatment of the same kind of phenomenon observed in Schwerdt (2000), but with a more conservative (almost Neogrammarian) conclusion.) Aryaman (☼) 15:46, 29 November 2007 (UTC)

Thanks, Pfold, this is very useful, and in my opinion confirms that we've got it about right when we say that the term has a narrower and a broader definition. We focus first on the narrower definition, then give details of the things which can also be involved under the broader definition. I'm sure there are lots of details we can improve, but the overall shape of our article is a fair representation of what you have just given as the range of viewpoints. The question is whether some synthesis of this can be added to the article, perhaps in a very short paragraph (since few readers will want to know exactly which scholars have said what), perhaps at the end of the section "general description". (Actually, given that the paragraph already mentions the two views, it would be enough to have a sentence: "scholars who use the broader definition include A, B, C; those who prefer the narrower definition include X, Y, Z.) Will you add this?

Aryaman, your information is also very interesting. Clearly, we don't do original research, and we don't run with fashion, so the overall shape of the article should continue to reflect the established majority view. A section further down on alternative perspectives, however, would be very welcome. Why don't you write this? --Doric Loon (talk) 14:32, 29 November 2007 (UTC)

Oh, I didn't mean the reference to Schwerdt to imply that there is anything wrong with the article. Her view is certainly not mainstream, and is only now beginning to appear referenced in other literature. I'm not even sure if - at this early stage - it could be used to justify a section like the one you propose (granted, there may be others of like mind, but I have not run across them yet). IMO, a footnote linked to one of the relevant passages from the section on Gerneral Description (e.g. the one dealing with the areal nature of the individual changes) should suffice without offending WP:UNDUE. I was thinking something like:
Schwerdt (2000) has argued that the name 'High German consonant shift' is misleading and perhaps even inappropriate, as it does not adequately reflect the areal discrepancies of the individual changes underwent by the effected West Germanic dialects.
The details on her book:
  • Schwerdt, Judith (2000). Die 2. Lautverschiebung: Wege zu ihrer Erforschung. Heidelberg: Carl Winter. ISBN 3-82-531018-3. 
Of course, feel free to edit as you see fit. Also, I can provide some quotes if anyone is interested in expanding it. Aryaman (☼) 15:46, 29 November 2007 (UTC)

OK, I've added three footnotes, one as Aryaman suggests, and two noting the results of Pfold's Census. Probably that will do for now, though in time we might give greater prominence to these things. But let's wait and see how future discussions emerge. Meanwhile, however, Pfold has a job to do: all the scholars whose names I have copied from your census into the footnotes need to be properly included in the bibliography. Can you do that please? --Doric Loon (talk) 22:53, 29 November 2007 (UTC)


I would like to see the old map again. I mean the map where the Low Saxon, Low Franconian and other dialects outside Germany were not edited out. The dialect continuum does not stop at borders. However, File:Benrath-Speyer.PNG got deleted. -- Zz (talk) 22:02, 3 May 2008 (UTC)

incorrect translations[edit]

In case anyone notices that I deleted some of the examples and changed one, the reason is because some of the translations were wrong and the correct translation wouldn't be an example.-- (talk) 01:27, 1 June 2008 (UTC)

This is precisely what makes Wikipedia the greatest encyclopedia on earth: anyone can display their brilliance! :) —Aryaman (talk) 02:01, 1 June 2008 (UTC)

Sorry, but I will have to revert those changes. Those were NOT translations, they were cognates. German Zeit means time, but it is cognate with tide. Zeit:tide is therefore a pair which illustrates the z:t parallel. (Zeit:time would NOT illustrate it, because they are unrelated words.) Likewise cup and Kopf illustrate p:pf. It is entirely irrelevant for the sound shifts that subsequent semantic shifts have caused the pairs to have divergent meanings. Kopf has enjoyed a fascinating semantic development: first, it meant cup, then it was used jokingly for the top of a bald head (I've heard English dome used that way too), then it became the standard word for head. French tête has a similarly slangy etymology, from a Latin word meaning roofing tile. So sure, you cannot now translate cup with Kopf (as you could in the 13th century). But that is not what THIS article is about.--Doric Loon (talk) 09:15, 1 June 2008 (UTC)

Actually, the article already had a note above the table, and a footnote at the bottom, which carefully explain all this.--Doric Loon (talk) 09:21, 1 June 2008 (UTC)

Perhaps it should be mentioned that caput/capitis means "head" and cuppa/cuppae means "cup" in latin and that the word could likely be a romance substrate word in germanic languages. The word "kopf" only appears by the year 790 in Old High German texts (Reichenauer Bibel-Glossar) and every information given to the time before is linguistic reconstruction, hence a theory. --El bes (talk) 12:28, 2 June 2008 (UTC)
According to Kluge, Lat. cūpa < cuppa is the substrate, and was an early loan (surely prior to the 8th century; probably into (N)WGmc). According to the Oxford Dictionary of English Etymology, there were two distinct words in OE, namely cuppe "cup" and copp "head", which then experienced some blending in form, with "cup" winning out over "head" semantically. All interesting stuff. But is a note really necessary here? —Aryaman (talk) 13:26, 2 June 2008 (UTC)

Careful with substrat theory, please, as it's quite controversial - though I rather like it. I don't think it is helpful to want to give the entire history of a word; the point here is purely that a form kop existed, changed to kopf, and exists in the parallel language as cup, thus displaying p:pf. The only reason for giving more info at all is to stop people with a smattering of German and no understanding of the topic changing "cup" to "cap" (as we once had) or deleting it because they don't know what to do with it. This sort of thing has happend four or five times now and is becoming wearing. What I have now done is to insert a link from Kopf and Zeit to Wiktionary. If you want to give more detail on the words themselves, put it there. --Doric Loon (talk) 13:56, 2 June 2008 (UTC)

Be also careful with your example, because the unshifted form "kop" is a linguistic reconstruction and is not found in the early 8th century texts, especially not in the Alemanic and Bavarian ones, where the soundshift is supposed to have originated. --El bes (talk) 06:46, 4 June 2008 (UTC)

Quite right: I should have asterisked *kop. But it is not in the article text anyway. --Doric Loon (talk) 13:03, 4 June 2008 (UTC)

In-line references, GA status[edit]

This article shows an apparent lack of in-line citations. While I'd fancy that most info provided here is easy to retrieve from the literature given, the matter is somewhat awkward. One particular point is also that Rheinischer Fächer which is given as a reference is insufficiently referenced itself. I'm unsure whether a Good article reassessment might be appropriate here, but if someone could address this issue (which should be easy enough), it would certainly improve the article. G Purevdorj (talk) 09:35, 7 June 2009 (UTC)

GA Reassessment[edit]

This discussion is transcluded from Talk:High German consonant shift/GA1. The edit link for this section can be used to add comments to the reassessment.

I'm not comfortable with initiating this GA review because I feel (though I am no expert on anything Indo-Germanic and did not verify the infos) that it meets GA requirements. However, there are two exceptions: The guidelines (2a) and (2b) are not met.

(a) it provides references to all sources of information in the section(s) dedicated to the attribution of these sources according to the guide to layout;
(b) it provides in-line citations from reliable sources for direct quotations, statistics, published opinion, counter-intuitive or controversial statements that are challenged or likely to be challenged

There is probably little controversial information in this article, but the violation of (2a) is pretty significant. I hope that someone familiar with this stuff might be ready to improve this, else this article must be demoted.

In the course of reading this article, I also came upon a few other points that, however, are not crucial to this GA review:

  • “High German experienced the shift /sp/, /st/, /sk/ → /ʃp/, /ʃt/, /ʃ/ in initial position.” word-initial position?
  • The exact nature of gemination should be addressed, eg why do you write “*k→hh” when we have modern High German /x/? Or “*p→ff”: couldn’t we determine a rule that in order for {IPA|/ɪ/} to stay short, “/f/” has to be geminated in {IPA|/ʃɪf/}, thus relegating the gemination to a synchronic process at the interface of phonology and phonotactics? I’m not saying that my suggestion must make sense, but it appears quite obvious, so that it might be discussed and rejected.
  • The use of standard orthography only doesn’t make understanding easier. IPA might sometimes be helpful.
  • Standard High German for <Pferd> seems to be {IPA|[fɛɐt]} (or something like that), not {IPA|[pfɛɐt]}. <pf> seems like mere orthography to me if it doesn’t coincide with a syllable boundary. I'm pretty sure this could be verified with appropriate literature.

G Purevdorj (talk) 23:19, 18 June 2009 (UTC)

Hi. A couple of fair points there. But <pf> really is pronounced /pf/ except in some dialects. Initial position means word-initial position. (We distinguish initial, medial and final position - it's standard terminology.) I thought we did use IPA quite a lot in this article, and you have cited some of it.
The only controversial things in this article are the question of whether soundshifts outside the core group are included as belonging to the shift, the question of whether to divide the core group into two or three phases, some aspects of the dating, and the theory about Gothic. All of those have been given good citations already. Most of the rest could be taken from ANY of the standard reference books, and I am not sure how helpful it would be to cite the same pages again and again in each section. Does good article status really require this when the whole article can be based on a single source?
--Doric Loon (talk) 23:47, 18 June 2009 (UTC)

As long as I don’t verify my claim about /pf/, my perception isn’t of any relevance. Being from Hannover, I perceive myself as saying /kamf/, not /kampf/, /ferd/ instead of /pfert/ etc. But I’ll try to check this during the next few days from my fellow speakers. Be this as it may, back to in-line citations and references:

  • Vennemann, Ernst Schwartz aren’t cited. The same holds for several books cited in the footnotes.
  • Primary sources such as Sachsenspiegel, Deutschenspiegel and other sources are not cited (it is not improbable that some of these texts are part of more complex editions).
  • “If, as some scholars believe, Lombardic was an East Germanic language”. Scholars aren’t identified.

The problem that any of these infos could be retrieved from almost any standard source is problematic for me as well, but as far as I understand the GA requirements, yes, the article would probably better confirm to GA requirements if at least every major paragraph had a reference. So I do think that referencing is an issue for this article from this more general and the more specific point of view mentioned first. G Purevdorj (talk) 20:33, 19 June 2009 (UTC)

Aha, Hanover. A friend from Lüneburg says the same.
It would be very easy to give a citation for every paragraph; the reason we didn't do it is because it would be the same citation for every paragraph, and because it would be arbitrary, picking one of the hundred possible books for no particular reason. But if you want it, we will do it. I don't have time for that just now, but I will offer to find editions for the Deutschspiegel and Sachsenspiegel. The point about scholars who think Lombardic is East Germanic is more difficult. I think we took that from the Wiki article on Lombardic, so I suppose we would need to copy over a citation from there. --Doric Loon (talk) 08:22, 20 June 2009 (UTC)
Well, take your time (I guess we are not talking of months, but of weeks) - the article itself is really good (congrats, by the way), so it may remain on hold for quite a while. G Purevdorj (talk) 10:52, 20 June 2009 (UTC)

A bit of time has passed, but the source issues have not been addressed so far. Just as a reminder. G Purevdorj (talk) 13:08, 10 August 2009 (UTC)

Originally I intended to do something about the source situation of this article myself if nothing happens, but I won't find the time. If no one declares that s/he will do something about it within a week, or if no considerable improvements are made within three weeks, I will delist this article. G Purevdorj (talk) 13:23, 23 September 2009 (UTC)
I'm not happy to do so as I think that this article is within close reach of the GA criteria, but as no improvements were made during the considerable time span of this reassessment, I close it by degrading this article. G Purevdorj (talk) 18:42, 7 October 2009 (UTC)

Silly reverts[edit]

What is the correct procedure when somebody breeches the revert rules? --Doric Loon (talk) 21:49, 15 August 2009 (UTC)

If I got it correctly, any admin could block a user that breaks the revert rules for a specified time. You would have to turn to an admin, then. Anyway, the events are a bit distant for an immediate blocking. Thus, I guess, you'll have to revert again. G Purevdorj (talk) 12:12, 17 August 2009 (UTC)
You report violations of WP:3RR at WP:AN/EW.
You report unfriendly and insulting edit summaries (e.g., [2] and [3]) at WP:WQA.
Then you the rest of you can go to WT:MOS, where you'll find some grammatically capable editors. I have no doubt that they will tell you that Drphilharmonic is correct with at least the majority of his changes. For example, where, which is supposed to indicate a location, is not a grammatically correct alternative for the phrase in which or the word that.
Drphilharmonic is wrong to impose his personal preference for that instead of who, which is exactly as wrong replacing British spelling with American spelling. However, this set of changes moves the article from one grammatically correct version to another grammatically correct version. If that weren't an acceptable choice for referring to humans, then Shakespeare would have written "He jests at scars that who never felt a wound." WhatamIdoing (talk) 19:28, 17 August 2009 (UTC)
Then he isn't correct with the majority of his changes, because the majority of his changes involve "which-hunting" (and "who"-hunting), as well as stamping out perfectly good adverbs like "evidently" and "intervocalically". He doesn't seem to believe that "while" can mean "whereas", though even the highly prescriptive AHD acknowledges that meaning. As for using "that" in reference to humans, of course it's possible sometimes, but not in the sentences he changed, where "who" sounds natural and "that" sounds atrocious. I think he ought to focus more on adding content to Wikipedia and leave the copy-editing to the native English speakers. +Angr 20:09, 17 August 2009 (UTC)
The actual issue here is not important enough to make me want to draw in Admins, but that kind of ill-considered linguistic prescription is certainly a disturbance of the peace, and this gentleman's edit history (also on many other articles) suggests that making relative pronouns conform to a minority practice is something of a crusade for him, which really is not good. What bothers me more is the discourtesy in the edit summaries, which rings alarm bells. --Doric Loon (talk) 22:27, 17 August 2009 (UTC)
His grammar is correct in his which-hunts; his conduct (his unilateral decision to impose his preference over the objections of other editors) is undesirable. And, yes, Doric, many editors, including myself, have repeatedly encouraged him to apply the principles behind WP:RETAIN to this situation. WhatamIdoing (talk) 00:26, 18 August 2009 (UTC)
Of course his conduct is the main problem, but his edits themselves, even if not always strictly speaking wrong, often have the effect of making the text harder to read and comprehend because the prose sounds so unnatural. +Angr 05:41, 18 August 2009 (UTC)
As far as what is "correct" is concerned, see English relative clauses#That and which. --Doric Loon (talk) 06:46, 18 August 2009 (UTC)

Peculiar Choice of "Low German" sample dialect[edit]

The so-called Low German words cited as examples in Phases 2, 3, and 4 are all examples from the Plautdietsch dialect, also known as "Mennonite Low German", rather than from the mainstream North Low Saxon of the lower Weser and Elbe districts and Holstein. It is rather like using Robbie Burns's Lowlands Scottish as the source of English word examples, rather than using standard English words - or quoting Schwabisch (or even "Pennsylvanian Dutch") words as examples of "High German", instead of using modern Standard German words. It just strikes me as a bizarre choice - why Plautdietsch (which is not actually common in northern Germany)? In any event, Plautdietsch words have undergone their own vowel and consonant changes which muddy the waters: e.g. k > kj, and r > a (between a vowel and consonant), and long a > o, so that Low German Wark > Plautdietsch Woakj. But the Low German Wark is a better comparison with Dutch werk and High Alemannic Werch, than the Plautdietsch Woakj which has obscured the Low German vowels and consonants with later changes.

Here are standard Low German equivalents* to your Plautdietsch ones:

Phase 2: Appel (vs Aupel), scharp (vs schoap), Katt (vs Kaut), tamm (vs tom), Wark (vs Woakj), Sparling (vs Spoalinkj), Nacht (the same), tru (vs trü).

Phase 3: doon (vs doonen), Moder/Mudder (vs Mutta), root (the same), bidden (vs beeden) NB: Low German beden corresponds to High German beten, not to bitten.

Phase 4: dat (vs daut), denken (vs dinken), döstig/dörstig (vs darstijch), Broder (vs Brooda), du (vs dü).

/v/->/b/: Leef/Leev (vs Leew), half (vs haulf), Lebber/Lewer (vs Läwa), sülv-/sülb- (vs self). NB: see Low German Salf/Salv for HG Salbe. Low German tends to have f,v, or b for HG b, depending on word position. /b/ is usually found before /n/ e.g. sülben/sülm vs sülvst (cf HG selben, selbst), and sometimes before /r/ e.g. över/öber vs HG über. /f/ occurs when the consonant becomes final through loss of unaccented -e e.g. MLG salve > NLG Salf/Salv.

all examples of Low German come from the Plattdeutsch-Hochdeutsches Wörterbuch, by Wolfgang Lindow, Institut für Niederdeutsche Sprache (Bremen, 1998), which is a much more mainstream source than Plautdietsch. —Preceding unsigned comment added by (talkcontribs) 09:03, 8 July 2010

When we wrote this article, we put in English, German and Dutch examples for everything. I think that is enough, as it covers all the variants relevant to the point of the article and illustrates them from the best-known standard languages. Later somebody added Low German. I think that clutters the article, without any additional benefit. After all, we could also include Yiddish or Swiss-German or Scots or any number of other language forms as additional examples of one side of this or the other, but I think that would only obscure the point of the article. Since the Low German forms are obviously questionable anyway, I would think seriously about dropping them altogether. --Doric Loon (talk) 12:38, 10 July 2010 (UTC)

Being a fan of Low German, I would love to see it represented here, but I agree with your statement that English and Dutch are standard languages and adequately illustrate non-shifted West Germanic dialects, in contrast with standard High German representing the shifted dialects. And, unlike Low German, they are much better known by the average, non-specialist reader. Delete away. —Preceding unsigned comment added by (talk) 15:15, 17 July 2010 (UTC)

Connection refused[edit]

Rhineland fan: The connection was refused when attempting to contact

(The link points to however.) Site badly constructed? Page redirected but requiring authentication? I can only make conjectures. — Tonymec (talk) 12:13, 26 August 2010 (UTC)

Terminal devoicing[edit]

The assertion here that the German word "Tag" is pronounced in the same way as the English word "tack" is incorrect. In English "tack" would have a short vowel sound followed by an unmistakable ejective 'k', whereas the German word "Tag" has a long vowel, followed by a far softer and lengthier 'gh' sound.Examinator (talk) 10:45, 16 March 2011 (UTC)

Maybe change it to "not unlike"? I think the comparison is still helpful. --Doric Loon (talk) 11:26, 16 March 2011 (UTC)
Perhaps 'talk' is a better match? CodeCat (talk) 11:51, 16 March 2011 (UTC)
Depends what kind of English you speak. I am from Scotland. For me, English tack and German Tag are as close as words in different languages ever can be. Talk is quite different. But the point is, if you are going to have interlanguage comparisons of sounds, you can take it as read that you are always dealing with approximations. In this case, it is not the vowel which is relevant anyway, but the final consonant. I think that to say German Tag sounds more like tack than tag is as good an illustration as you can give for an English speaker who wants the point put briefly. --Doric Loon (talk) 09:14, 5 July 2011 (UTC)

Katze and Ratte[edit]

Both of these words have -tt- in other Germanic languages (as well as Latin), and yet one of them shifted and the other didn't. Why is that? CodeCat (talk) 19:42, 11 April 2012 (UTC)

Ratte (like English rat) is borrowed from Vulgar Latin or an early Romance language, so presumably the borrowing post-dates the productive period of the shift? --Doric Loon (talk) 20:39, 11 April 2012 (UTC)
But is that really the case? From what sources I can find (Dutch etymology dictionaries), neither word has a conclusive etymology. CodeCat (talk) 00:25, 12 April 2012 (UTC)
Ratte and Ratz ~ Ratze (Bavarian) are native to German. This word is dealt with in some detail in Guus Kroonen's The Proto-Germanic n-Stems, pp. 221-2 (2008; [4]). Kroonen reconstructs the word as an n-stem (PGmc *raþō, gen. ruttaz < PIE nom. *Hrót-ēn, gen. Hrt-n-ós), which then gave rise to multiple variations; thus, *raþþ- > OHG radda ~ ratta > G Ratte; *rattōn > OHG ratza > G Ratz(e); *rad- > OHG rato > MHG rate. In other words, both are shifted and native to German and no borrowing took place. Torvalu4 (talk) 18:04, 13 November 2012 (UTC)

/sk/ > /ʃ/[edit]

The article currently says 'occurred in most of West Germanic', but how much of that is really true? It occurred in Old English and in High German, but it did not occur in Dutch or Frisian. I don't know about Low German. So this should probably be rephrased. CodeCat (talk) 01:02, 5 September 2012 (UTC)

Low German "loans"[edit]

Since standard German is, and always was from the very beginning of standarization, a hybrid of elements from many High and Low German dialects, can it make sense to call a Low German form in standard German a "loan"? Surely this is of a fundamentally different nature from, say, a loan word from Latin? To treat it as the same seem to me to fail to understand for example that "Haven" was never transferred sideways from LG to HG, but rather was carried forward from LG into universal German. I know there is a political POV which wants to accentuate the degree to which Low German is a separate language (whatever that means) rather than a German dialect (the distinction seems unscientific) and recent changes in this article seem to me to have a whiff of that - though I am not suggesting other editors consciously have that agenda. They may just be confused by the two different meanings of Hochdeutsch. At any rate, none of my text books call LG forms "loans", and I would be looking for citations from linguists if this is to be kept. Any thoughts? --Doric Loon (talk) 14:04, 10 December 2012 (UTC)

If I understand you correctly, you don't dispute that Hafen is ultimately from LG. However, I don't understand your point about being "transferred sideways" vs. "carried forward" (what do these expressions mean?): either the word is from LG or it isn't. If it is, it's a loan; if it isn't, it's not a loan. Technically, the dialect/language distinction is immaterial for something to be a loan ([5]; [6]) so that doublets may arise, e.g., Eng ditch vs. dike. LG and stand. Germ are not dialects of each other, but rather distinct languages, and have been considered as such since their earliest attested precursors (OHG vs. Old Saxon). Admittedly, non-linguists don't necessarily acknowledge this. Germ. isn't a "hybrid", if by that you mean something like a creole/pidgin that had equal input from 2 source langs. If you mean its vocabulary, it too isn't any more "hybrid" than most other languages. For more, read specialist literature, like The Germanic Languages, by Koenig. In any event, Pfeifer's etym. dict. (2005 in print [7]; another in Dutch [8]) shows Hafen to be from LG. Torvalu4 (talk) 05:30, 12 December 2012 (UTC)
Hafen is from LG. That's not in dispute. Haben is from HG. Both have come into Standard German, which is a mixture of LG and HG. My point is that since standard German never was pure HG (despite Germans using the word Hochdeutsch loosely in a way that confuses everyone) but rather, ALWAYS had LG forms in it for as long as the standard language existed, you cannot speak of a loan taking place. I think you need to go to those books you are citing and look yourself. If you can find a linguist wriitng that Hafen is a loan word, I'll accept I'm wrong, but I don't think you will find that. I'm not arguing with you about the other details - just about the term. I think lingusts don't use the term "Lehnwort" for LG forms in German. If you want to prove me wrong, find a citation. --Doric Loon (talk) 20:10, 12 December 2012 (UTC)
You argue that because standard German takes forms from different areas of Germany, therefore Hafen can't be considered a loan but merely another word incorporated into the standard language. I'm not sure if you can look at it that way. Modern standard German is based on east central German, an area considered High German, and most of its forms match up at least as far as the consonant shift goes. As with all standard languages, forms from different areas are taken in, but that in itself doesn't mean Hafen can't be a loan. It could have been loaned into one of the otherwise High German-speaking areas that the standard was based on. After all, a (large) harbour isn't something you would typically find inland in High German territory, so the word must have seen more frequent usage in the coastal areas - Low German speaking areas. It is also telling that there are very few words that don't conform to the consonant shift as it is in east central German; that implies that the vast majority of standard German vocabulary originates there. After all, the scribes of the time didn't go around travelling to look for words; they just used words that were taken to be common and well understood. That's what "standard" is all about; it's a compromise, an average of usage. So that makes any words that do not come from the core area of the standardisation automatically stand out. I believe that it's likely that east central German didn't even have a word for harbour, and if it did, that it was pretty rare, and therefore easily supplanted by a Low German loanword. CodeCat (talk) 00:57, 13 December 2012 (UTC)

Map problem[edit]

Most of the article goes back a very long time, however the map only shows current situation. German was spoken along the Baltic coast to north of Memel (Klaipeda in Lithuania today) , entire Silesia (upper and lower) and there were islands of German speaking territories even 1000's of km into Russia. Boeing720 (talk) 23:24, 11 July 2014 (UTC)

Boeing720, you are entirely right. Nevertheless, the map was correct in so far as it went, and was more helpful than having nothing at all. I therefore would like to put it back until something better is found - adding a comment that it refers to the modern situation. But if you would like to replace it or complement it with a map of the 8th-century situation, I would be extremely happy about that. --Doric Loon (talk) 22:22, 10 November 2014 (UTC)
The problem with choosing an earlier map is that one for the OHG period can only be based on a very thin corpus of locatable texts from a handful of individual monasteries, which therefore cannot show true isoglosses, while a map for the MHG period or later will show large areas of settlement in the east where the patterns of shifted consonants reflect post sound-shift migration rather than the distribution of the original sound change, though of maps for the last 100 years or so at least have the virtue of showing genuine isoglosses based on surveys. Not sure what the solution is!--Pfold (talk) 13:28, 12 November 2014 (UTC)

External links modified[edit]

Hello fellow Wikipedians,

I have just added archive links to one external link on High German consonant shift. Please take a moment to review my edit. If necessary, add {{cbignore}} after the link to keep me from modifying it. Alternatively, you can add {{nobots|deny=InternetArchiveBot}} to keep me off the page altogether. I made the following changes:

When you have finished reviewing my changes, please set the checked parameter below to true to let others know.

You may set the |checked=, on this template, to true or failed to let other editors know you reviewed the change. If you find any errors, please use the tools below to fix them or call an editor by setting |needhelp= to your help request.

  • If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
  • If you found an error with any archives or the URLs themselves, you can fix them with this tool.

If you are unable to use these tools, you may set |needhelp=<your help request> on this template to request help from an experienced user. Please include details about your problem, to help other editors.

Cheers.—cyberbot IITalk to my owner:Online 10:49, 11 January 2016 (UTC)

Move this article?[edit]

Oddly, the opening sentence of this article gives no indication that in fact the most common term in English is not consonant shift but sound shift.

In the standard works on my shelves or my hard disk I can find only four authors who use the term consonant shift - Waterman, Wells, Ringe, and Prokosch. Against that are ten for sound shift - Wright, Keller, Nielsen, Russ, Young & Gloning, Barbour & Stevenson, Voyles, Salmons, and König & Auwera. If we cut out works over 50 years old, it becomes 2 against 9.

At the very least, the opening sentence needs rewriting, but surely, unless people can find a lot more handbooks that use consonant shift, this article should be moved. --Pfold (talk) 17:50, 23 March 2017 (UTC)