European Languages and Scoring on the SAE Phonology Test

[Wattmann]

I have dropped this, as people obviously cannot appreciate the effort that went into this (there is no detailed online repository of dialectisms of Icelandic, so I pieced them together from videos)

[merijn]

Why are you limiting yourself to phonology?If I go to the Wikipedia article on SAE most features mentioned there have to do with syntax and morphology.

[Wattmann]

Because morphology is not shared to that extent in a sprachbund (see: Finnish, see: Hungarian, see: Basque, see: Wakashan and Salish, see: Armenian) and is a primary feature of a lingustic family of languages - Icelandic, being North-Germanic and in a special position, would share all points with Faroese, most with Western Norwegian, and plenty more with Swedish and Danish (descending amount as one goes south)

[merijn]

While that is true to some extent, there are quite a few features mentioned on that wikipedia page that are shared for instance between Germanic and Romance, without going back to PIE, so there is certainly an areal influence. Form the first list on the page this is the case at least for features 1, 3 , 5 and 11, and for most others I am not sure.

[Wattmann]

Those features aren't shared by Swedish, Norwegian, Faroese and Icelandic, though I don't know much about Danish's status. They are not as obvious as the phonological shared features alveolar: fricatives and affricates, uvular approximant (not shared with English), front rounded vowels (Gothic, Vandalic, Proto-Norse didn't have any front rounded vowels, English lost them and certain dialects (example: Kiwi ) re-innovated them back).

Non-English Germanic languages have "Dative + Copula.3.SG + Adjective" for experiences (it is nice to me = I feel nice; it is cold to me = I feel cold), and languages influenced by continental Germanic have no pronoun dropping, while there are also features that German(ic) has(have), French lacks and Slavic languages have.

Many of those shared features (such as external possessors) are probably from direct language influence (Portugese is no neighbour of Germanic), or possibly substrate influence (Visigoths spoke Gothic in Iberia till the 8th century AFAIK), not inter-linguistical exchange.

Of course, sprachbunds share some morphology (Balkan sprachbund and its definiteness craze, centred around Bulgarian and Slavic Macedon), but their phonological simmilarities are far more apparent (central vowels, palatal frenzy, clicks as answers to yes-no questions), and their lexical borrowing is also very obvious (Hungarian <as(ov)>, <gjuvecs>(?), Slavic <vojnik> <*berza> <da> <pusti(ti)> - I feel wiktionary is a lying son of a Dornish whore, but no matter).

The point is, languages are obviously in a sprachbund if they share morphology, but it requires prolonged exposure and is easily confused with ancestral sprachbunds, super/substrate influence or plain parallel evolution or common origin.

[merijn]

Those features aren't shared by Swedish, Norwegian, Faroese and Icelandic, though I don't know much about Danish's status.

All of the features I mentioned as not coming from PIE are shared by North-Germanic. You could argue that the suffixal definite article doesn't count, but the most peculiar thing about SAE is that the languages mark the distinction between definite and indefinite. Also you could argue that feature 11 doesn't count for all North-Germanic languages, because some of them don't have subject agreement, but it is true for those that have subject agreement. For the features I didn't mention, I think that North-Germanic has most of the features listed on the Wikipedia page, but I am no expert on North -Germanic

Non-English Germanic languages have "Dative + Copula.3.SG + Adjective" for experiences (it is nice to me = I feel nice; it is cold to me = I feel cold), and languages influenced by continental Germanic have no pronoun dropping, while there are also features that German(ic) has(have), French lacks and Slavic languages have.

And your point is?

Many of those shared features (such as external possessors) are probably from direct language influence (Portugese is no neighbour of Germanic), or possibly substrate influence (Visigoths spoke Gothic in Iberia till the 8th century AFAIK), not inter-linguistical exchange.

I don't understand what you mean here. What is the difference between "inter-linguistical exchange" and "direct language influence"? And can you point me to some people who have argued that SAE's grammatical features are the result of a substrate influence?

The point is, languages are obviously in a sprachbund if they share morphology, but it requires prolonged exposure and is easily confused with ancestral sprachbunds, super/substrate influence or plain parallel evolution or common origin.

The thing is, the languages participating in SAE, have fairly to extremely well-described histories, so people can easily rule out "ancestral sprachbunds" and "common origin"

[Herr Dunkel]

The thing is, the languages participating in SAE, have fairly to extremely well-described histories, so people can easily rule out "ancestral sprachbunds" and "common origin"

Bullshit. You're thinking of German, French and English there. What about Frisian? Elfdalian? Gotlandish? Or Kashubian, Polabian, Sorb? Occitan? Romansch? Norman? Pars pro toto, my man.
Also, "fairly" is relative. Abui is also fairly well known in the world.

[merijn]

The thing is, the languages participating in SAE, have fairly to extremely well-described histories, so people can easily rule out "ancestral sprachbunds" and "common origin"

Bullshit. You're thinking of German, French and English there. What about Frisian? Elfdalian? Gotlandish? Or Kashubian, Polabian, Sorb? Occitan? Romansch? Norman? Pars pro toto, my man.
Also, "fairly" is relative. Abui is also fairly well known in the world.

I don't really see how this is relevant to the discussion. I claim that we can rule out a common origin of say the perfect with have + passive participle. We can rule that out because we know the history of Germanic and Romance very well, probably better than any other language family. We know that it wasn't in proto-Romance, and we also have a pretty good idea that it wasn't in proto-Germanic.
And even the knowledge of Old-Frisian or Old-Occitan dwarves what we know of the history of most other languages in the world, and other lesser known Germanic and Romance languages where there may be no historical record, we still know a lot about their history because we can compare them to very close relatives of which we know the history. We know where they came from, we know what sound changes they have undergone. I'd say, even if there are no historical records of Elfdalian (I couldn't tell you) we know much more about its history than almost any language of sub-saharan Africa or the Americas.

[Radagast revived]

Because morphology is not shared to that extent in a sprachbund (see: Finnish, see: Hungarian, see: Basque, see: Wakashan and Salish, see: Armenian) and is a primary feature of a lingustic family of languages - Icelandic, being North-Germanic and in a special position, would share all points with Faroese, most with Western Norwegian, and plenty more with Swedish and Danish (descending amount as one goes south)

This is not true. Morphological features are every bit as prone to diffusion as phonology. There are not a single phonological feature characterizing the Mesoamerican sprachbund for example. Anyway Whorf, who defined the concept of SAE was explicitly not talking about shared phonology but about grammatical categories.

[Herr Dunkel]

The thing is, the languages participating in SAE, have fairly to extremely well-described histories, so people can easily rule out "ancestral sprachbunds" and "common origin"

Bullshit. You're thinking of German, French and English there. What about Frisian? Elfdalian? Gotlandish? Or Kashubian, Polabian, Sorb? Occitan? Romansch? Norman? Pars pro toto, my man.
Also, "fairly" is relative. Abui is also fairly well known in the world.

I don't really see how this is relevant to the discussion. I claim that we can rule out a common origin of say the perfect with have + passive participle. We can rule that out because we know the history of Germanic and Romance very well, probably better than any other language family. We know that it wasn't in proto-Romance, and we also have a pretty good idea that it wasn't in proto-Germanic.
And even the knowledge of Old-Frisian or Old-Occitan dwarves what we know of the history of most other languages in the world, and other lesser known Germanic and Romance languages where there may be no historical record, we still know a lot about their history because we can compare them to very close relatives of which we know the history. We know where they came from, we know what sound changes they have undergone. I'd say, even if there are no historical records of Elfdalian (I couldn't tell you) we know much more about its history than almost any language of sub-saharan Africa or the Americas.

I'm not psychic, you know?
You said "we know a whole lot about every SAE language", and I'm immediately supposed to asume "have + participle" and that knowledge of Elfdalian is bigger than any subsaharan language (no it's not. Elfdalian is a dialect of Swedish and was until recently written in runes - it has a shallower known history than, say, Maya, or Naahuatl, or Amharic, or Afrikaans)

This is not true. Morphological features are every bit as prone to diffusion as phonology. There are not a single phonological feature characterizing the Mesoamerican sprachbund for example. Anyway Whorf, who defined the concept of SAE was explicitly not talking about shared phonology but about grammatical categories.

I seriously doubt this. If this were truth, we'd see a crapload of monoconsonantal roots in Egyptian Arabic, and ergativity in Armenian; Mongolian would be monosyllabic, and Manchu would lose cases.
Grammar is much harder to diffuse - an example is Anglo-Norman influence on Old English, or even Old East Norse on Old English - loanwords import new instances of rare or previously uncontrasting phonemes (Anglo-Norman alveopalatals, Old Norse dentals).
Old Norse contributed quite a lot to the phonology of English, and that's now known as Scots - one thing stuck in English from Old Norse, though, and that's "they/their".
Although, Piraha borrowed its entire pronoun system from a neighbouring language, so it's not unheard of. I cannot say with certainty that phonology diffuses with more ease (grammatical features are often imported by substrate influence - Celts and Angles), but if you're so certain that morphology is as prone to speading as phonology, do back it up, shan't you?

[Astraios]

If this were truth, we'd see a crapload of monoconsonantal roots in Egyptian Arabic, and ergativity in Armenian; Mongolian would be monosyllabic, and Manchu would lose cases.

Who told you that's what'd happen? Are you an expert already? The rest of your post is just as ridiculous.

[Whimemsz]

Yeah, Darkgamma, you don't know shit about any of this. Stop setting yourself up as a Linguistics Expert. Accept that you're a noob and try to learn more: before making a bunch of wrong assumptions and then presenting them here as facts, go read about how linguistic areas and diffusion work; you'll end up making less of a fool of yourself and looking less like an Eddy.

NE: I'm certainly no expert on this subject myself, though I am interested in it. Anyway, no one is denying that phonological features are frequently shared by neighboring languages; but it's simply not true that grammatical features are much more rarely shared. Radagast already did give you an example of a linguistic area defined solely by lexical and morphosyntactic traits (Mesoamerica) rather than phonological ones. I think it's reasonably uncontroversial to say that there's a hierarchy of features that are more to less easily dissimulated across language boundaries IN GENERAL: lexemes > calques > phonological stuff > morphosyntax. But it absolutely does not follow that morphosyntax is much less likely to diffuse than phonology; it just tends to require more prolonged and intimate contact between groups. And in the end these are all tendencies.

[merijn]

The thing is, the languages participating in SAE, have fairly to extremely well-described histories, so people can easily rule out "ancestral sprachbunds" and "common origin"

Bullshit. You're thinking of German, French and English there. What about Frisian? Elfdalian? Gotlandish? Or Kashubian, Polabian, Sorb? Occitan? Romansch? Norman? Pars pro toto, my man.
Also, "fairly" is relative. Abui is also fairly well known in the world.

I don't really see how this is relevant to the discussion. I claim that we can rule out a common origin of say the perfect with have + passive participle. We can rule that out because we know the history of Germanic and Romance very well, probably better than any other language family. We know that it wasn't in proto-Romance, and we also have a pretty good idea that it wasn't in proto-Germanic.
And even the knowledge of Old-Frisian or Old-Occitan dwarves what we know of the history of most other languages in the world, and other lesser known Germanic and Romance languages where there may be no historical record, we still know a lot about their history because we can compare them to very close relatives of which we know the history. We know where they came from, we know what sound changes they have undergone. I'd say, even if there are no historical records of Elfdalian (I couldn't tell you) we know much more about its history than almost any language of sub-saharan Africa or the Americas.

I'm not psychic, you know?
You said "we know a whole lot about every SAE language", and I'm immediately supposed to asume "have + participle" and that knowledge of Elfdalian is bigger than any subsaharan language (no it's not. Elfdalian is a dialect of Swedish and was until recently written in runes - it has a shallower known history than, say, Maya, or Naahuatl, or Amharic, or Afrikaans)

You may not be psychic, but I think it is safe to assume that you have reading skills and that you can see I am arguing that we can rule out a common origin because we know quite a lot of the origin of these languages. And in a way Elfdalian doesn't have a shallow history. Its history goes back all the way to PIE and we can make reasonably good good guesses how it evolved from PIE to till now, at least much better guesses than on how Swahili for instance evolved out of Proto-Bantu, let alone how it evolved out of Proto-Niger-Congo (if there is such a thing as Proto-Niger-Congo). In other words, although there may not be that much actual records, we do know its history better than the majority of other languages in the world.
(I have to admit though that I forgot about the Meso-American and South-American civilizations, and that Semitic is spoken in Ethiopia. However, Other than those languages I Icannot think of a language in the Americas and in Sub-Saharan Africa that we know as much of its history than any Germanic language.

[Acid Badger]

I seriously doubt this. If this were truth, we'd see a crapload of monoconsonantal roots in Egyptian Arabic, and ergativity in Armenian; Mongolian would be monosyllabic, and Manchu would lose cases.

Since you're mentioning Armenian, modern Eastern Armenian does show some grammatical features that Classical Armenian didn't have and that have similarities in non-related neighbouring languages. There's for example the Dative case that, sometimes, is used for the direct object, like in Georgian. Or analytic verb constructions by means of a participle and an auxiliary, which look a lot like what is going on in Nakh-Daghestanian. Or the two past tenses with the aorist encoding witnessed and the perfect unwitnessed events. Or certain agglutination-like features in it's nouns and verbs and so on.

Oh and most indigenous languages of the Caucasus do have and ergative case, but what that case is actually used for varies heavily among the language families and even the more related languages.

Also what Whimemsz said.

[Nortaneous]

wow this is a stupid argument. stupid but informative, like most arguments here are.

i've got wattmann on ignore and i can't be bothered to click the thing to read his posts but i'm guessing it's limited to phonology because that's what that test that somehow keeps going around is about, and the test is about that because standard conlang morphology/syntax are much less typically european* than standard conlang phonology, and also i know fuckall about either so i can't get angry about them

* finnish is not particularly SAE. finnish plus split-S is even less so.

[zompist]

That Wikipedia article gives an incorrect impression, I think, of how much Whorf's "standard average European" term is used. I just checked three books on historical linguistics and none of them mention it, though two of them talk about a European or West European linguistic area.

(Whorf's own uses of the idea are pretty suspect anyway; he didn't define or explore it in anywhere near the detail of Haspelmath. It was mostly his bogeyman for contrasting with his claims about the worldview of Amerindians. One of his few explicit claims about SAE-- that it grammaticalizes past/present/future-- is contradicted by his own fricking language.)

[Herr Dunkel]

more to less easily dissimulated across language boundaries IN GENERAL: lexemes > calques > phonological stuff > morphosyntax. But it absolutely does not follow that morphosyntax is much less likely to diffuse than phonology; it just tends to require more prolonged and intimate contact between groups. And in the end these are all tendencies.

That's what Wattman said without the coherence XD

This devolved pretty fucking quickly...

[Wattmann]

i've got wattmann on ignore and i can't be bothered to click the thing to read his posts but i'm guessing it's limited to phonology because that's what that test that somehow keeps going around is about, and the test is about that because standard conlang morphology/syntax are much less typically european* than standard conlang phonology, and also i know fuckall about either so i can't get angry about them

* finnish is not particularly SAE. finnish plus split-S is even less so.

First of all, that' very rude of you.
Second, I'm trying to see how SAE the phonologies are - I will change the title if you're so bothered by it.

[brandrinn]

OK, if people want to talk about the borrowability of morphology, I guess that's a worthwhile topic. But keep in mind the SAE test we devised was supposed to be about phonology. That was the point from the beginning. It's not something Wattmann made up.

Is anyone still interested in the thread's original purpose, or should we rename the thread?

[Wattmann]

OK, if people want to talk about the borrowability of morphology, I guess that's a worthwhile topic. But keep in mind the SAE test we devised was supposed to be about phonology. That was the point from the beginning. It's not something Wattmann made up.

Is anyone still interested in the thread's original purpose, or should we rename the thread?

Thank you, brandrinn, but I think this is far too much. It is also what happened the last time I wanted to make a contribution. I'm dropping this.

[Salmoneus]

Because morphology is not shared to that extent in a sprachbund (see: Finnish, see: Hungarian, see: Basque, see: Wakashan and Salish, see: Armenian) and is a primary feature of a lingustic family of languages - Icelandic, being North-Germanic and in a special position, would share all points with Faroese, most with Western Norwegian, and plenty more with Swedish and Danish (descending amount as one goes south)

This is not true. Morphological features are every bit as prone to diffusion as phonology. There are not a single phonological feature characterizing the Mesoamerican sprachbund for example. Anyway Whorf, who defined the concept of SAE was explicitly not talking about shared phonology but about grammatical categories.

Yay! Rada! Welcome back!

[Nortaneous]

Is anyone still interested in the thread's original purpose, or should we rename the thread?

In the unlikely event that I ever get around to finishing that series of posts, I'll probably just split it off into a new thread, so.

Although I guess I should at least write something about vowels, since I'm going to be fucking with that a lot in the Hathic langs...

[Radagast revived]

This is not true. Morphological features are every bit as prone to diffusion as phonology. There are not a single phonological feature characterizing the Mesoamerican sprachbund for example. Anyway Whorf, who defined the concept of SAE was explicitly not talking about shared phonology but about grammatical categories.

I seriously doubt this. If this were truth, we'd see a crapload of monoconsonantal roots in Egyptian Arabic, and ergativity in Armenian; Mongolian would be monosyllabic, and Manchu would lose cases.
Grammar is much harder to diffuse - an example is Anglo-Norman influence on Old English, or even Old East Norse on Old English - loanwords import new instances of rare or previously uncontrasting phonemes (Anglo-Norman alveopalatals, Old Norse dentals).
Old Norse contributed quite a lot to the phonology of English, and that's now known as Scots - one thing stuck in English from Old Norse, though, and that's "they/their".
Although, Piraha borrowed its entire pronoun system from a neighbouring language, so it's not unheard of. I cannot say with certainty that phonology diffuses with more ease (grammatical features are often imported by substrate influence - Celts and Angles), but if you're so certain that morphology is as prone to speading as phonology, do back it up, shan't you?

The languages you mention are not located in diffusion zones, they are not members of any prachbunder that I know of. And you just mention some random changes that would have happened if that were the case which is of course a non-sensical way of falsifying a positive claim through counterfactuals.

Try reading Kaufman & Thomason "Language contact and creolization" or Thomason "Language contact". They quite explicitly say that everything is borrowable and that the only determinig factor is degree of contact. There is a huge literature on syntactic borrowing that you could take a look at as well. The Balkan sprachbund also has several morpholocial traits.

In certain way phonemes are harder to borrow because they tend to require full native competence. Highly marked phonemes are unlikely to be borrowed unless the population is fully bilingual or there is a complete language shift, whereas syntax can be borrowed from high prestige to low prestige languages even with sporadic bilingualism if the contact is intense.

[Radagast revived]

That Wikipedia article gives an incorrect impression, I think, of how much Whorf's "standard average European" term is used. I just checked three books on historical linguistics and none of them mention it, though two of them talk about a European or West European linguistic area.

(Whorf's own uses of the idea are pretty suspect anyway; he didn't define or explore it in anywhere near the detail of Haspelmath. It was mostly his bogeyman for contrasting with his claims about the worldview of Amerindians. One of his few explicit claims about SAE-- that it grammaticalizes past/present/future-- is contradicted by his own fricking language.)

The statement about Whorf's claim being refuted by English is a strawman promulgated by Chomsky. Whorf does not claim that English has a future tense (which it actually kind of does just not morphologically marked - but why would you expect that from a mostly analytical language anyway?). He says that English (and SAE) conceptualizes time as divided into past, present and future which it undeniably does. But yes that wikipedia article makes it look as if SAE is a standard concept in linguistics.

[zompist]

(Whorf's own uses of the idea are pretty suspect anyway; he didn't define or explore it in anywhere near the detail of Haspelmath. It was mostly his bogeyman for contrasting with his claims about the worldview of Amerindians. One of his few explicit claims about SAE-- that it grammaticalizes past/present/future-- is contradicted by his own fricking language.)

The statement about Whorf's claim being refuted by English is a strawman promulgated by Chomsky. Whorf does not claim that English has a future tense

Yes, he does. "The three-tense system of SAE verbs colors all our thinking about time." (LTR p. 143) And the rest of the passage makes it clear that he is not confusing tense with time, as his purpose is precisely to distinguish them.

English doesn't have a future tense. It has several modal constructions that indicate future time among other things. Whorf's view of "SAE" verbs is very old-fashioned, with little apparent awareness of how aspect and modality work across languages, or even how much variation there is within Europe.