Back in 2007 I posted about Justin Rye’s brilliant discussion of proposals for spelling reform, adding that I was doing so thanks to a comment by David Marjanović; now, once again thanks to a comment by David Marjanović (in this thread), I’m posting about Rye’s discussion of “What would English be like if 1066 hadn’t happened?” It’s a lot of fun (except for the inhabitants of London, who get nuked in 1983), and I recommend it to anyone who enjoys playing with sound changes and dialects.
Comments
Trackbacks
-
[…] Hat explores Ingalric, an alternate-historical English […]
Umm, say what? “Demolition of spelling reform”? Demolition of arguments against spelling reform, mostly, with a concession at the end that the political issue really is a show-stopper. German: four countries, 15 years, widespread discontent. English: 50 countries, probably 50 years if not 500 years. “The great thing about standards is that there are so many to choose from — and if you don’t like any of them, wait till next year.”
A just as intriguing question would be: what if the Normans in just 100 years hadn’t changed their language from Norse to French?
Otherwise it is, from my point of view striking how similar the English and the Scandinavian grammars are. The differences I spontaneoulsy come to think of are: (1) the use of the particle do/did in questions and negative clauses; (2) the use of present participle (-ing) as a marker of imperfective aspect; (3) the position of the adverb in subordinate clauses (I see, that he not is here, would be the Swedish word order); (4) the strict ‘V2-rule’ in Swedish, i.e. the verb is always in second place in a sentence: ‘Today bought I a car’.
Even in subtle things like the aspectual difference between worked and have worked is the same in Scandinavian as in English (while it’s a little different in German).
As for vocabulary the basic English one already is Gmc. Were it not for the Normans, the English perhaps would have said ‘berg’ and ‘dale’ instead of ’mountain’ and ‘valley’, ‘weighty’ instead of ‘important’, ‘lean’ and ‘sproak’ instead of ‘salary’ and ‘language’ and ‘wittenship’ instead of ‘science’. They would also maybe have been less restrictive in the use of compounds. So ‘linguistics’ might have been ‘sproakwittenship’.
No, what’s really ‘alien’ about English in a Gmc. context is the phonology. From Iceland to Austria I see nothing strange in their pronunciation, although deviating from my native tounge. But it is something very special about English with its eg. diphtongs not found elsewhere. Don’t ask me why, if the Normans are to blame or if it is an internal matter of things due to insular isolation (‘Fog over the English channel – the continent is isolated!’).
JC: You’re right, of course, “demolition” was a bad choice, and I’ve changed it to “discussion.” But he ends by saying “In case you’re wondering, no, I don’t believe that this sort of wholesale spelling reform would be a workable proposition, but I’m so sick of watching Aunt Sally reform proposals being pelted with ridiculously inadequate arguments that I thought it would make a nice change if I wrote something equally biassed and unfair in the other direction,” so the takeaway is that, whatever the theoretical arguments might be, spelling reform ain’t gonna happen. Which is fine with me.
From Iceland to Austria I see nothing strange in their pronunciation, although deviating from my native tounge. But it is something very special about English with its eg. diphtongs not found elsewhere.
Not that English isn’t weird, but what about Danish?
And for diphtongs, I give you Scanian:
And Bavarian:
Oh. And Scanian vowel phonology still looks suspiciously similar to the IE system of laryngeals and grades to me.
Scanian and Bavarian? There are of course dialects all over the world which are more or less incomprehensible even by their closest neighbours. I was talking about standard Scandinavian, Dutch, English Icelandic and German.
And when it comes to the Danes, haven’t you Norwegian oil trillionaires already made enough fun of them? After all they gave us Kirkegaard, HC Andersen and Niels Bohr.
There is no such thing as mocking the Danes “enough”. As if the princesses weren’t bad enough, what other culture practices kartoffelbegravelse?
About Scania (Skåne), the southernmost part of contemporary Sweden, I believe it’s dialect to be a real shibboleth. This fertile plain land was transferred from the Copenhagen king to the Stockholm one in the treaty of Roskilde in 1658. Through the decades to follow there were attempts from both the emperors to claim the sovereignity of the area and anyone who spoke in the Danish dialect could be killed by the Stockholm king and vice versa. So the Scanians came up with inventing ‘Scanian’, neither Danish nor Swedish, in order to survive.
The most obvious feature of Scanian is its diphtongs (and even triphtongs), found neither in Danish nor in Swedish. From Danish it has inherited the vocalisation of stops word finally or intermedially. So Swedish kaka (cake, bisquit) is in Scanian kaga (Danish kage), köpa (buy, cf. ‘cheap’) is köba (Danish købe) and gata (street, cf. gate) is in Scanian gada (Danish gade). Otherwise all characteristics of modern Scanian are Swedish.
@Stefan Holm – don’t be so quick to dismiss Bavarian as a mere “dialect”. It is spoken in a continuum from Munich to Vienna and has arguably more than 13 million native speakers, which means it is more widely spoken than any individual Scandinavian tongue
don’t be so quick to dismiss Bavarian as a mere “dialect”. It is spoken in a continuum from Munich to Vienna and has arguably more than 13 million native speakers.
Catalan has even more speakers and no one takes that seriously as a language in its own right.
A question for German speakers and Germanists: Standard German … Bruder [bruːdɐ], Brüder [bryːdɐ]
I see that the final unstressed er is transcribed as a low central vowel [ɐ], which is what I have heard from (apparently Standard) German speakers, but the sequence er when stressed (as in Erde ‘earth’) does not use the same vowel but a diphthong. I am interested in this pronunciation and transcription because in one branch of the language family I study, which was first described by Franz Boas more than a century ago, there is a high central-to-back unrounded vowel, occurring under stress in many words, usually transcribed as [ɯ], as in the independent pronoun [n’ɯ:yu] ‘I, me’.
Boas (a German speaker writing before the IPA became standard) wrote er for [ɯ:] , and I have been wondering how he himself pronounced the German sequence, since he claimed “I hear distinctly r” in the language variety in question, a sound that I definitely have never heard there. Based also on internal and comparative criteria it is not possible that the language in question or its proto-language ancestor ever had a sound corresponding to any one of the many varieties of “r”.
des: Catalan has even more speakers [than Bavarian] and no one takes that seriously as a language in its own right.
The Catalans certainly consider it a language.
If it is not a language in its own right, what is it a dialect of? I find it very similar to Occitan at least as much as to Spanish.
Don’t mind des, he’s a jokester.
@des:
I assume you are joking.
Almost certainly Boas’s German was non-rhotic (a development of the 19C, I think), so he pronounced coda /r/ as [ɐ̯] and coda /er/ as [ɐ]. That’s a lot more open than [ɯ], but I won’t say there’s no resemblance at all.
I assume you are joking.
Good lord, of course he’s joking — I already said so. No need to have gone to all that trouble.
Thanks, JC, but the problem is not the unstressed er coda (as in Bruder) but the stressed er as in German Erde which Boas heard in words like [n’ɯ:yu] ‘I, me’ or [n’ɯ:n] ‘you’. In slow speech the vowel may be lightly diphthongized as [ɯ:ɐ] but this sequence never sounds like the stressed er in Erde which starts with a mid front vowel as I have always heard it.
I agree that there is a resemblance between [ɯ] and [ɐ] since they are both central vowels, one high and the other low, but they are not allophones of each other, and neither resembles the mid front vowel.
And then, I suppose, there are northern-influenced subvarieties of the southern varieties of Nothern Bavarian and down we go to individual villages.
Getting back to Ingalric. Somehow physicists resist the urge to predict the outcome of the Brownian motion, why linguists can not?
Entschuldingung, Vanya. Es war nicht meine Meinung, ein deutsches dielekt zu reduzieren. Ich habe nur von die Sprache von Goethe, (hochdeutsch) ‘Über allen Gipfeln ist Ruh, kaum einen Hauh spürest Du ..’ gesprecht.
Stressed coda er would be [ɛɐ̯].
To be fair, much of the discontent could have been avoided had the reform not been done in such an undemocratic way. The governments and nebulous, almost anonymous experts got together, came up with a plan, and then it was implemented. Minor changes were made several times just before and during the implementation period, alleviating some of the less well thought-out issues but likely adding to the confusion… which remains substantial.
To also be fair, the reform was tiny compared to anything or almost anything I’ve ever seen suggested for English.
It’s not “a little different” but completely absent in German – but that’s an innovation shared with French, where the passé simple and the passé composé have likewise merged their meanings.
English compound nouns, especially but not only in writing, can actually grow to intimidating sizes and incorporate whole phrases. This is just hidden by the orthography which insists on spaces between the parts.
I’m intrigued. Please explain! 🙂
That’s pretty extreme backpedalling you’re doing here. 🙂
On the other hand, the inverted diphthongs – which I didn’t even know about – are specifically North Bavarian.
I thought “only” 7 million?
This clearly is what Boas meant when he wrote he “heard ‘r’ in there”.
Maybe what he was aiming at by writing er was [əɐ̯]; this is not a sound that occurs in German As I Know It*, but I suppose [ɛɐ̯] is the closest German has to offer… From a German perspective, it takes some practice to learn to distinguish between [ə], [ɤ] and [ɯ]**, so maybe Boas was aiming at “some strangled un-vowel forming a diphthong with [ɐ̯] in careful pronunciation”.
* I’m trying to say that neither my knowledge of German dialectology nor even of Standard German accents is complete enough to rule such things out with certainty.
** …though [ɤ] and the English [ɜ] are also perceived as similar to [œ].
English compound nouns, especially but not only in writing, can actually grow to intimidating sizes and incorporate whole phrases.
“Constantinople Bagpipe Manufacturing Company”, for example, or “Danube Steamship Company captain”. Or worse yet, “salt-and-pepper-bearded mother-of-pearl-handled-pistol-wearing rent-a-cop.”
Scanian has developed most vowels into diphtongs. Most of these diphtongs start as schwa-like central vowels and end up as semi-vowels or approximants. Short, unstressed vowels keep the original realisation (or close enough). I’m not able to describe it in phonetic detail, so I turned to the ‘Net. Here‘s a somewhat idealised recent description meant as basis for an orthography. There’s generally a three-way split between
Long stressed vowel: Central wowel + long semi-vowel/approximant
Short stressed vowel: Central vowel + short semi-vowel/approximant
Short unstressed vowel: Plain vowel.
In short, I think we see something similar to IE full, reduced and zero grades. The o-grade is not there, but the grammatical environment is not the same — and note that unstressed is colored by a preceding vowel.
The Scanian vowel inventory is larger than reconstructed for PIE, but some of the phonemes are quite similar and may be prone to reduction (if not by by real-world development, so by incomplete reconstruction in 6000 years).
Trond, are Skåne and Skandinavia cognate?
AJP: It’s uncertain. Scandinavia is actually the result of an ancient typo: it should have been Scadinavia. The latter part means ‘island’, and the OED agrees that the first part is cognate with Skåne, but other sources point to a word for ‘damaged, imperfect’.
The latter part means ‘island’
Our ‘øy’ < *aɣʷija-
I don't think there's agreement on what the placename denoted at the time before it found its way into international cartography. It may have been the whole region of Skåne or just Skanør, the old marketplace by the Sound.
German Schaden “damage, detriment”? So that Skaþin-awija would be Damage Island?
…I could sort of see that name being used for the rocky coast of Norway, I guess, which will probably damage your ship if you aren’t very careful.
BTW, German Au(e) refers, nowadays, to riverside forests that are flooded every spring, so “coast” is probably not too far form “island”.
Trond: Our ‘øy’ < *aɣʷija-
David: German Au(e) refers, nowadays, to riverside forests that are flooded every spring
Are these words related to Latin aqua ?
(I misremembered the reconstruction, I always do. *aɣʷjó:- is more like it. And that ɣʷ is overly precise, since its phonemically gʷ.)
m-l: Are these words related to Latin aqua ?
Yes. as a derivation. The cognate is å “river” < *áhʷo:-. It sounds exactly like its Fr. cognate <i<eau.
Imperfect, I think, in the sense that it isn’t actually an island. But nobody in the Latin-speaking world could have known that at the time, so I’m skeptical. Me, I like the idea of the “Perilous Shore” better. English has only a few akwa-words, like ait/eyot ‘river island’ and the first vowel of island (the “s” in the spelling comes from the etymologizing spelling of isle < insula.
Aqua and its cognates. They’re limited to Italic, Germanic, and possibly Celtic.
Or of course the other way around. 🙂
Aqua and its cognates. They’re limited to Italic, Germanic, and possibly Celtic.
And English: ewer
That’s borrowed from French.
Imperfect, I think, in the sense that it isn’t actually an island. But nobody in the Latin-speaking world could have known that at the time, so I’m skeptical.
Also, I don’t think Germanic speakers started using the øy word to name islands of some size until much later — and certainly long after the initial settlement: Just in the neigbourhood we have Gotland, Öland (yes, I know), Sjælland. Lolland, Langeland. (Irland and Ísland are both larger as islands and younger as names.). But Bornholm for some reason — maybe its steep coasts. Given the semantic range of the cognates it must even have denoted flooded banks and the like before it was extended to any landmass surrounded by water. With that in mind, the fact that Skanør is a low peninsula, so low that it’s slowly disappearing below sea level, makes it a likely example of an early *aɣʷjó:-.
Nevertheless, I’m irrationally fond of the idea that Skaði refers to the whole Scandinavian peninsula, personified by the eponymous giantess/ancient godess Skaði, the daughter of the sea giant Þjazi, whose name is suspiciously close to Sami čáhci “water”. At the time of recorded mythology, Skaði was a minor godess of the wilderness, hunting and skis, unhappily married to the sea god Njǫrðr, but there’s toponymic evidence for a cult to her, and the earls of Hálogaland, Northern Norway, traced their lineage back to Sæmingr, a son of her and Óðinn (or Freyr, depending on the sources).
Or of course the other way around.
It’s probably simpler to describe the conditions for [ɣʷ] as a conditioned allophone of gʷ than the other way around. But there’s something more to this. [ɣʷ] is both the voiced allophone of hʷ and the fricative allophone of gʷ. Or maybe it’s an either-or, depending on the sequence of Grimm-Verner sound shifts.
I haven’t mentioned German Eiland yet; that’s poetic for “island” – the normal word is Insel.
Ooh, can of worms. 🙂
What’s easier may depend on which pronunciation was used in initial position. For the unrounded *g, North Germanic points toward [g], West Germanic points toward [ɣ] (found today in southern Dutch; shifted to /j/, not /d͡ʒ/, before front vowels in English; shifted to [g̊], not [k], in High German), and East Germanic probably points nowhere because Wulfila used the same letter for both allophones and no further information seems to have survived. If a plosive was used in initial position, we can say “fricative allophone between vowels”; if a fricative was used there, we can say “plosive allophone behind nasals”…
About Spanish I’d say “plosive utterance-initially and behind nasals, approximant elsewhere”.
This ended when (sometime after Verner and Kluge had run through) stress ceased to be phonemic in some ancestor of Proto-Germanic.
David, what do you mean, phonetically, by “[g̊], not [k]”?
He means by [g̊] a voiceless but lenis consonant, as opposed to [k] which is voiceless and fortis. You can hear the difference by whispering the English words gab and scab: the whispering suppresses voicing distinctions and the leading /s/ suppresses aspiration distinctions.
Thanks, JC. So it’s just a difference in the burst intensity? Is it the same in High German?
It’s pretty much the same throughout the Germanic languages. In English, aspiration, voicelessness, and fortis enunciation are pretty much aligned (except after /s/).
I think so, and I think the higher intensity is due to higher pressure from the lungs (not in the throat or somewhere). There’s very little research on this, because very few languages have a fortis-lenis distinction that isn’t reinforced by another distinction that is easier to investigate (usually voice and/or aspiration). I have to run right about now, but I have a paper I’ll cite.
Found the paper.
David is probably right in presuming it has to do with the amount of expiratory air. There is a pretty well known experiment, where you record people pronouncing words like spin, still and skate. Then you edit the recording by cutting off the initial ‘s’, play the edited version in front of native speakers and ask them to write down, what they here. They won’t, as maybe expected, write pin, till and Kate, but instead been, dill and gate.
The experiment can be done in most Germanic languages with the same outcome. It’s the aspiration vs lack of it, that counts – and that’s the real difference between ’till’ on one hand and ‘dill’ or ‘still’ on the other.
Yes. But High (in the sense of Southern) German is the exception: it doesn’t do aspiration at all, but does do fortis/lenis. Aspiration is about voice onset timing: English pin is the result of not starting to voice the [ɪ] immediately after the [p] is released.
You’re of course right about the ‘voice onset timing’. But either it’s combined with or results in an increased air flow. Hold the lower end of a, say, letter size paper (or even the palm of your hand) close to your lips and pronounce words beginning sp-, st-. sk- or b-, d-, g- and then compare with words beginning with p-, t-. k-.
I know about High German and I also suspect that things are more or less opaque in Danish due to their very own varieties of stops (the glottal ones). But I would be surprised, if any significant difference were to be found between English and Swedish. My paper (size A4 – we don’t use letter) certainly will flutter after a single, initial voiceless stop.
Unfortunately I only took one year of French during my school years. But I remember our teacher trying to get us to pronounce calcul: Say it like in Swedish ‘kalkyl’ (borrowed from French of course) he told us, but try not to breathe – then you will get the ‘k’:s right :-).
Getting rid of aspiration and voicing was what the High German consonant shift was all about. 🙂
…Unsurprisingly, most High German dialects don’t even make the fortis/lenis distinction anymore. The Swiss have merged it with the length distinction (all fortes, in some dialects even at the beginnings of words, have become long lenes), eastern Austria has dropped the distinction, a wide area from Swabian to East Central German (like what’s nowadays called Saxon) has done the same (called binnenhochdeutsche Konsonantenschwächung “Inner German consonant weakening”), and those Central Bavarian dialects that connect the latter two areas, like mine, have generally lost the /tː/-/t/-/d/ distinction between vowels and at the ends of words – the word-final merger is carried over into Austrian Standard German, so that Tod “death” and tot “dead” are homophones like in northern Germany, but in the opposite way and for the opposite reason. 🙂
We were taught to do this in one of the first English lessons, because aspiration doesn’t come naturally to us.
Uh, apparently binnendeutsch is preferred nowadays (-hoch- redirects here), making my translation correct.
I have always supposed, probably from Alsatian, that the fortes-lenis distinction is just consonant length. Didn’t know that there are other ways to make that distinction.
Tod “death” and tot “dead” are homophones like in northern Germany, but in the opposite way and for the opposite reason.
Don’t tell me you are taking after the Danes, who back in time merged voiceless stops with voiced between vowels and word finally. C.f. Swedish/Danish: kaka/kage (cookie); gata/gade (street); köpa/købe (buy). Will the next step be a similar reduction of short consonants making most of them inaudible for a non native? Will Tod, tot and Tor all become [to]? 🙂
That’s what it has merged with in Not Too Low Alemannic; and it’s also, unfortunately, what the terms “fortis” & “lenis” were originally invented for. :-/
That would require voicing, which is unthinkable. 🙂
Perhaps not so unthinkable. $EMPLOYER is an ex-startup, now acquired, that does automated analysis of legal documents (mostly contracts) to find things like termination provisions, jurisdiction, and so on. It’s much cheaper than paying lawyers to do it, especially when you have thousands of pages of documents to look at in a M & A, though lawyers still have to review the results.
Though most of our customers care about English and Japanese, we are expanding to other languages, and I have been working on German, as comments scattered across the blog will indicate. But now an actual German-speaking lawyer is needed to review my work, and so we brought in Peter, a consultant. His American accent is so sharp that I thought he was a native, but no, he’s from Vienna-not-Virginia, and has “only” been here fifteen years. In particular, I had him review the outputs of my decompounding algorithm, so we have talked about a lot of German words and phrases, though of course in an English matrix.
Anyway, and ObHat, I have noticed two things: one, how rhotic my pronunciation is (inherited from my mother, as I have also said), with [ʀ] or at least [ʁ] for almost every postvocalic /r/, with the exception of the occasional [ɐ] for final -er; and two, how his final “voiced” stops are indeed voiced, which surprises me very much.
I missed that. And geographically I wouldn’t expect that from your mother either – too far east, I’d have thought.
Either that’s an English accent, or you’re merely hearing the lack of final fortition – the “voiced” stops are voiceless, but lenis.
My mother’s accent was much closer to stage pronunciation than to anything local. Indeed, she was forbidden to speak local dialect as all the other children in the village did, on pain of whippings from the aunts that were raising her (her father was in America and her mother was dead); she and they were the only family of bourgeois origin in the village, and Distinctions Must Be Preserved.
Looking at the WP.en description of stage pronunciation (henceforth BD), it seems clear that her accent (henceforth MC) was far more closely related to that with some loosening. In the order given there:
1) BD /r/ was [r] or [ɾ] until 1957, when [ʀ] was also permitted. MC /r/ is definitely [ʀ]; I’m not sure if my mother sometimes used [ʁ] or if that’s just me.
2-3) Neither BD nor MC have /r/-vocalization or syllabic consonants at all.
4) MC does not, unlike BD, have [ə] > [e] in final position in Greek and Latin loanwords.
5) MC has final fortition to the max, though I am not sure that my mother went so far as to fortition /s/ > fortis [z] in Absicht as BD demands; it sounds like lenis [z] varying to [s] when I say it.
6) I am unsure about BD’s hyper-aspiration of non-initial voiceless stops. That sounds like a particularly “stagey” feature and I certainly don’t have it, and I don’t think MC did either.
7) BD prescribes that a historically voiced obstruent after a historically voiceless one remains voiced, as in abdanken, where /p/ is voiceless and /d/ voiced. I’m not sure about MC’s accent; I basically do this, but have English-style unreliable voicing for all the historically voiced stops.
Two definitely non-stage features of the MC accent:
(a) A complete merger between long e and ä.
(b) The occasional appearance of [x] for /g/ in final position, something she deprecated in herself and told her students not to imitate.
I’ll listen more closely to Peter’s accent when he comes back in a week.
I didn’t even know BD was fully rhotic. It’s telling that the German article is very short and just about useless, talking about almost nothing but Siebs’s original intentions.
Full syllable-final fortition is probably universal everywhere north of the White-Sausage Equator.
Exaggerated aspiration is one of the deliberately artificial features optimized for stage use. Elsewhere, aspiration is never stronger than in English.
By the way, JBR has apparently relocated his homepage to a new address http://jbr.me.uk/.
Thanks very much! I’ve updated the links accordingly.
I don’t know where else to put this, but since there’s some discussion of German non-rhoticity here, this is where I’ll leave the link:
https://youtu.be/iEDvozbyUMQ?t=816
The sound of Helmuth von Moltke the Elder (b. 1800), recorded in 1889.
That’s amazing. And apparently there are recordings (using a different process) from 1860!
Fully non-rhotic and generally unremarkable for… I don’t know where exactly, but that exact accent is alive and well.