English as a North Germanic language?

[zompist]

Merci. Now I'm just a little confused why Mark was refuting an assertion no one had made, but no matter.

The clue is in the first line-- "quick summary for those who don't know". I don't assume everyone has read T&K and has the facts about Michif at their fingertips. As Anders mentioned mixtures, it's good to know something about the few languages that can be called mixtures (and why English isn't one of them).

[Karavinka]

I'm not convinced.

I'm not going to rehearse what's been said already on this thread. Instead, I want to question the comment that "Old English simply died out" -- as if there is a clear-cut discontinuity between "Old" and "Middle" English and as if these two are monolithic units.

Much of the Old English texts as we know are "standardized" into Late West Saxon. The West Saxon standard was naturally abandoned after the Norman conquest, and since we conventionally draw a line between Old/Middle English at the time of the conquest, if anyone wants to see if Old English actually died out or not should compare the language before and after the conquest. At least I think it's a reasonable assumption.

There are quite a few texts that show "transitional" stages between Old and Middle English, and - fortunately for us - they all show different dialects of the language. Ancrene Wisse from former Mercia, Ormulum from former Danelaw and Ayenbite of Inwyt from Kent. If their hypothesis is correct, than we'd expect at least Ormulum to be written in a 12th century Norse dialect (and the "Old" period of Norse is usually held to last until 14th century).

Orm, the author of Ormulum, developed an idiosyncratic orthography for his (boring) poetic works.. and he was very careful with the language. He used doubled consonant to consistently indicate that the preceding vowel is short. He used different letters for hard and soft "g". He eliminated prevalent allography, such as "eo" vs "e."

And by the way, he was a Dane from former Danelaw, writing in 12th century. So, we actually have an unusually good idea of a certain 12th vernacular, spoken by a Dane, in Danelaw, contemporary with those who produced Grey Goose Laws in Iceland.

Now, I'm boring you with all these details just to show you that you probably cannot dig up a text that is temporally and spatially and linguistically closer to whatever that was spoken in Danelaw. So, here's an extract. You can decide if this is Scandinavian or not.

Forrþrihht anan se time comm
þatt ure Drihhtin wollde
ben borenn i þiss middellærd
forr all mannkinne nede
he chæs himm sone kinnessmenn
all swillke summ he wollde
& whær he wollde borenn ben
he chæs all att hiss wille.

As soon as the time came
that our Lord wanted
to be born in this middle-earth
for the sake of all mankind,
at once he chose kinsmen for himself,
all just as he wanted,
and he decided that he would be born
exactly where he wished.

[hwhatting]

Media Lengua is mixed in an entirely different way from Mednyj Aleut or Michif. It went ahead and picked a phonology (Quechua), syntax (Quechua), morphology (have I mentioned Quechua?) and lexicon (Spanish). It's still weird that a language would replace all of it's lexicon in this way, with very little harm coming to it's grammar, but it's different from a language having two morphologies, two phonologies, and being split across lines of lexical class, like a pair of conjoined linguistic twins.

That sounds similar to what I've read about English Romani (which is Romani lexicon and English syntax & morphology) or Russian Fena (thieves' language - Russian syntax & morphology with a lexicon consisting of words loaned from Yiddish, Romani, plus some modified Russian words and some phantasy formations). Like many secret group languages, they use the grammar of the carrier language, which is the L1 of its speakers, and just insert lexicon from other sources. In the case of English Romani, the lexicon was borrowed from Romani probably when there were still L1 speakers of "real" Romani (i.e. Romani with Romani syntax & morphology); in the case of Fena, the lexicon was borrowed from various sources - the main goal is that it was meant to be incomprehensible to outsiders. I could imagine that Media lengua started as an attempt by some Qechua speakers to differentiate themselves against others by using higher-prestige Spanish lexicon.

[Salmoneus]

Actually, that's something that apparently happened a lot in the Andes - there are quite a few languages that have been ichneumonised, as it were, by spanish.

Iirc at least one shows an intermediate stage: the grammar is native and the nouns are native, and theoretically all the verbs are native... but the native verbs (which have been reduced to just a handful) only act as auxilliaries, and the main verbs (well, not syntactically verbs, but that carry the verbal semantics) are all borrowed from Spanish. Apparently this is a common process - it's easier to borrow words in uninflected form, or in a set inflection, so in largescale borrowings you can just borrow, as it were, infinitives (which may not have been the true infinitive form in the donor language, of course) and stick all the messy morphology stuff onto the native auxilliaries.

[Ambrisio]

We can just as well argue that English is Sino-Tibetan, because its grammar is much closer to Mandarin Chinese than, say, Latin or German.

[TzirTzi]

Must English be either West- or North-? Can't it just be said to be a mix?

Not according to the tree model. We don't really have a robust vocabulary for describing the various degrees of language influence which fall well short of the kind of mixing you find in languages like Michif or Media Lengua.

We may not have a good vocabulary to describe it, but the scholarly community agrees and has agreed for half a century that the tree model is drastically misleading for the vast majority of scenarios of language change and divergence.

To take the WG/NG case:

In the early runic period we have a relatively homogeneous variety across what we'd later view as the North and West Germanic speech areas (albeit in a drastically sparse record). There is an interleaving of changes which apply to N+WG, NG, WG, and just parts of WG: these were clearly parts of the same extended speech community for a long time; some of the changes that distinguish them happen early in this time and some of the changes they share happen late in this time. Thus this was presumably a complex dialect continuum, rather like that found in German dialects today.

When population movements such as the AS migrations happen, they take a bunch of speakers from one point on the continuum and put them in contact with speakers on another point on the continuum. In the following few centuries, there are also some more drastic changes in the northerly parts of the NG continuum (I'm thinking of u-umlaut and syncope) which distinguish them more strongly, but even before NG speakers really start viking southwards, OE speakers are in contact with OF speakers are in contact with OSw speakers are in contact with ONorw speakers and innovations spread back and forth.

The result is that even before the viking invasions and the Danelaw:

• some of the features which we consider important, defining features of NG are found sporadically or in different form across bits of WG (e.g. sharpening in a couple of OHG words; OE and ON breaking are given unitary analysis by some schoalrs)
• some of the features which we consider important, defining features of WG are found sporadically across bits of NG (e.g. the irregularly syncoped forms of 'go' and 'stand' in OSw gā stā; j-gemination, of course, actually did happen sometimes in NG with velars, e.g. leggja cf. Got lagjan; traces of a past participle in *-ana-; occasional loss of unstressed R < *z)

tl;dr

The point I'm making is that the tree model is nonsense: languages can diverge a bit, converge a bit, borrow a bit, etc., in any sequence and almost without limitation. Saying that ME is either WG or NG is only ever an academic shorthand. The journalistic article quoted here seems to be entirely based on a very old-fashioned belief in the literal truth of the tree model.

NB. Faarlund is a respected and experienced historical linguist - I happen not to agree with many of his conclusions elsewhere as well, but nevertheless I'd be inclined to guess that this article is a misrepresentation of his thought rather than him just suddenly talking nonsense.

[dhok]

Media Lengua is mixed in an entirely different way from Mednyj Aleut or Michif. It went ahead and picked a phonology (Quechua), syntax (Quechua), morphology (have I mentioned Quechua?) and lexicon (Spanish). It's still weird that a language would replace all of it's lexicon in this way, with very little harm coming to it's grammar, but it's different from a language having two morphologies, two phonologies, and being split across lines of lexical class, like a pair of conjoined linguistic twins.

I think Media Lingua's track record with phonology is mixed- ie Quechua phonotactics are mostly respected but, for example, Media Lingua distinguishes phonemic voiced and voiceless stops, but Quechua does not.