A Reworking of German Language Classification
Bob
German expanded from 20 to 83 separate languages.
Admittedly, it's all a gigantic dialect chain pretty much, but you may
as well draw some lines somewhere.
grammatim
> http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
>
> German expanded from 20 to 83 separate languages.
>
> Admittedly, it's all a gigantic dialect chain pretty much, but you may
> as well draw some lines somewhere.
Not if such "lines" don't exist.
We all leared about the "Rhenish fan" from Bloomfield's textbook. If
there aren't bundles of isoglosses, what are you dividing?
benl...@ihug.co.nz
> http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
>
>
> Admittedly, it's all a gigantic dialect chain pretty much, but you may
> as well draw some lines somewhere.
If you like drawing lines, I guess.
I'd like to know what your intelligibility statements are based on.
Ross Clark
Harlan Messinger
Why?
Can you clarify something? In your response to the first comment, your
reply, "This is one of the lies of the sociolinguists," without it being
clear which part of the first comment you were referring to. Your target
seems to be the proposition that "the German dialect chain extent into
the Netherlands and include Dutch". You even say, "It is these languages
that these liars are referring to when they speak of this wonderful
dialect chain between German and Dutch." But then you say, "The truth is
that there *is* a dialect chain starting in Belgium and going to
Austria." So, there is a dialect chain, or it's a lie?
Bob
<hmessinger.removet...@comcast.net> wrote:
> Bob wrote:
> >http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
>
> > German expanded from 20 to 83 separate languages.
>
> > Admittedly, it's all a gigantic dialect chain pretty much, but you may
> > as well draw some lines somewhere.
>
> Why?
>
> Can you clarify something? In your response to the first comment, your
> reply, "This is one of the lies of the sociolinguists," without it being
> clear which part of the first comment you were referring to. Your target
> seems to be the proposition that "the German dialect chain extent into
> the Netherlands and include Dutch". You even say, "It is these languages
> that these liars are referring to when they speak of this wonderful
> dialect chain between German and Dutch." But then you say, "The truth is
> that there *is* a dialect chain starting in Belgium and going to
> Austria." So, there is a dialect chain, or it's a lie?
Hello. It is not true that near the German border, "German" and
"Dutch" speakers can talk to each other. They aren't really speaking
the German or Dutch language. They are speaking separate languages
like Limburgs, Low Dietsch, Southeast Limburgs, Bergish, Ripaurian
languages, etc. Those are separate languages, not "German" or "Dutch".
That's why they are lying.
There is a dialect chain, but if you read Ethnologue, they seem to
have chopped up an awful lot of these dialect chains into separate
languages. I bet 25% of the languages in Ethnologue are dialect
chains. Why do you have to draw the line somewhere? Because you do.
Because we have decided that Czech and Slovak are separate languages.
It doesn't make sense to say that everyone from Belgium to Vienna is
all speaking one language. Forget that.
Bob
It's a pilot study. The purpose is to stimulate further discussion,
research, analysis and especially criticism.
Some of it was done by analyzing and comparing texts, some was done by
reading which languages speakers say they can hear and which they
cannot. Further, most of the *major* dialect groups of German were
simply split off because most reviews of German linguistics say flat
out that they aren't intelligible with each other or with St German.
I left some intact. I wasn't able to prove Thuringian is
unintelligible with St German, so it stayed. And Central and Northern
Bavarian seem intelligible.
The rest is based on Ethnologue, standard German dialectology and
speaker statements.
For intelligibility to split off a language, I want 90%. Below 90% =
language, above 90% = dialect. Since Afrikaans and Dutch are ~80%
intelligible, it makes sense to split them.
This treatment is actually quite conservative. Speakers of Ripaurian
regularly speak of 100-150 *separate Ripaurian languages*, where
speakers can understand their own town and next one over or so, but
beyond that, not so much. There have been 150 *separate dictionaries*
already made for Ripaurian lects, all of which differ greatly in
lexicon, phonology, etc. But until I get hard evidence for
intelligibility, I won't split those off.
There are many dialects of Limburgs, but speakers say they are all
intelligible, so there's no splits.
Further, my understanding is that there are 20-70 major Swiss German
lect, and *many* of them are not mutually intelligible. But I need
more evidence than that.
If you go here:
http://www.lowlands-l.net/anniversary/index.php
You can see many mostly Low Germanic lects. You can pretty much figure
out which ones are dialects and which must be separate languages just
by looking at them.
benl...@ihug.co.nz
I see here:
1) "people say...speaker statements..."
2) "80% ... 90 %..."
3) "just by looking"
So (1) you talk to large numbers of people about dialect differences?
(2) You (or somebody) has applied some standard quantitative test to
many pairs of dialects?
(3) Some of it is just impressions?
Ross Clark
Harlan Messinger
Uh huh.
So say you have "lects" A, B, C, D, E, F, G, ..., Z, and each one is 90%
intelligible with the one on either side. And say you decide that A, B,
C, D, and E one language with C at the core and F, G, H, I, and J form
another language with H at the core, and you note that C and H aren't,
practically speaking, mutually intelligible. In what way is that
incompatible with observing that speakers of E and F can understand each
other very well? You're talking about these things as though they're
mutually exclusive and as though the lines you're drawing are
particularly meaningful.
Bob
Exactly, this is how Ethnologue, and others, typically split many
languages. You ask speakers. Hey, can you understand those people over
there? "Sure." Dialects. "Are you kidding?" Separate languages.
Something in between? Much trickier, but often speakers will attempt
to clarify via percentages how much they understand. Once they start
saying, "We find it really hard to understand those people," you're
almost always dealing with 2 languages. At 90%+, people usually say
they can understand them well, more or less, etc. When they say things
like, "We can understand them partly but not fully," you are often
dealing with ~60-70% intelligibility.
>
> 2) "80% ... 90 %..."
If you read through Ethnologue, they seem to be splitting these days
at 90%. And you read all the literature on intelligibility done by SIL
people, this is indeed what they say. They split at 89%.
>
> 3) "just by looking"
Not so difficult! Let us take a look here: Hamburgisch, Ollands and
Oldenburg, 3 Low Saxon lects:
Hamburgs:
http://www.lowlands-l.net/anniversary/hamborgsch.php
Ollands:
http://www.lowlands-l.net/anniversary/ollandsch.php
Oldenburg:
http://www.lowlands-l.net/anniversary/ollenborgsch.php
Quick observation shows us that Hamburgs and Ollands obviously must be
dialects of one tongue. Yet Oldenburg seems so different that it
doesn't seem possible for them to converse with the others at 90%+.
Keep in mind that much of science is simply observational, hunches,
intuition, etc. Sir William Jones famous discovery certainly was.
>
> So (1) you talk to large numbers of people about dialect differences?
> (2) You (or somebody) has applied some standard quantitative test to
> many pairs of dialects?
Yes, tests have been done and the results are that most German
"dialects" are in fact separate languages. Swabish, Bavarian, Swiss
German, Mainfrankisch, Pfalzisch, Luxembourgish, Hessian, Badisch all
have 40% intelligibility with Standard German.
Hutterite has 70% intelligibility with Pennsylvania German and 50%
with St German and Plautdietsch (Mennonite).
Problem is that to do this testing you usually need a funding source
and a team to go out and do it. I'd love to see lots more testing done
in German. In Chinese, too. Lots of places. Hardly anyone does this
for some reason, maybe because it's boring. Linguists are usually off
dabbling in highfalutin theory. Most testing nowadays is being done by
SIL by Bible translation purposes.
> (3) Some of it is just impressions?
Some, not much.
A lot also comes from the linguistic literature.
1X2Willows
>
> Yes, tests have been done and the results are that most German
> "dialects" are in fact separate languages. Swabish, Bavarian, Swiss
> German, Mainfrankisch, Pfalzisch, Luxembourgish, Hessian, Badisch all
> have 40% intelligibility with Standard German.
Better yet, there is no empiric "Swiss German" but once more, only
a plethora of dialects which may or may not be mutually intelligible.
Rule of thumb: The further apart in distance, the less commonalities.
This is the reason why in the majority of Cantons, St German is being
taught in school as a primary 'foreign language' from first grade on.
The rest teaches Italian and French respectively, because people
speak specific dialects of those in everyday life.
Schwiizertüütsch (Swiss German) is usually not written, except for
completely unregulated phonetic spellings, used by people sometimes
for fun or as an expression of identity; a trend which is most popular
among the younger generation nowadays.
If you have any further questions about SG, please ask away.
PaulJK
> On Mar 9, 7:58 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>> On Mar 10, 12:14 pm, Bob <lindsay.rob...@gmail.com> wrote:
>>> On Mar 9, 1:57 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>>>> On Mar 10, 5:51 am, Bob <lindsay.rob...@gmail.com> wrote:
>>>>> http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
[...]
>> I see here:
>> 1) "people say...speaker statements..."
>
> Exactly, this is how Ethnologue, and others, typically split many
> languages. You ask speakers. Hey, can you understand those people over
> there? "Sure." Dialects. "Are you kidding?" Separate languages.
> Something in between? Much trickier, but often speakers will attempt
> to clarify via percentages how much they understand.
It's laughably unrealistic to base that on such provably flimsy
value judgments.
You'd need "standard" (whatever that means) speakers
with IQ=100, all with the same understanding of what
"understand those people" mean, and with no prior exposure
to "those peoples'" language.
No prior exposure is needed since most people are
demonstrably unable to gauge how much exposure they
have had and to what extent it influences their value
judgments.
Take a well documented example of Czech and Slovak.
Around the world, thousands of Czechs and Slovaks meet
socially in private and in their various emigree clubs. They
speak in their respective national languages and dialects.
The first generation emigrees understand each other perfectly.
During the conversation they are often not even aware of
how different their languages really are. They all grew up
in a bilingual society. From early childhood they were
passively exposed to radio and television broadcasts in both
languages. Offical business was carried out in either language
throughout the whole federal republic. Typically the news
items were read by two presenters alternating between the
languages, one item in Cz, next in Sk, and so on. Bilingual
texts in both languages were often not used. For example,
money was printed in both languages but each denomination
was entirely printed either in Czech or Slovak.
When children of the emigrees were old enough to join
mixed groups of adults in conversation the parents realized
they had a problem. The kids didn't quite understand people
speaking other dialects a didn't understand at all the other
language. The same problem affecting young people exists
more recently in Czech Republic and Slovakia where a new
young generation has grown up since the split of the old federal
republic.
So, if you ask people over 40, "do you understand those
people", they'd swear blind there's nothing to it. But, if you
ask under 25s, they'll tell you they often have no idea
what the conversation in the other language is about, let
alone understand what exactly is being said.
Standard Czech and Slovak are quite clearly different languages.
These days, because the native speakers live in separate
countries, it's not difficult to find speakers of each who have
not been exposed much to the other language and the languages
prove to be indeed sufficiently different for the lack of
understanding to be noticeable.
But, how do you use this method to gauge objectively a mutual
level of understanding between people living in geographical
or social proximity to each other, in the same administrative
unit, and being exposed to the others' language without even
realizing the extent and effects of the exposure?
> Once they start
> saying, "We find it really hard to understand those people," you're
> almost always dealing with 2 languages. At 90%+, people usually say
> they can understand them well, more or less, etc. When they say things
> like, "We can understand them partly but not fully," you are often
> dealing with ~60-70% intelligibility.
That's like conducting an IQ test, where you don't make
people solve the puzzles and give you objective answers,
instead just tell you how easy they think each question is
on the scale of 0 to 100%. :-)))
pjk
benl...@ihug.co.nz
Statements of this kind are notoriously influenced by attitudes to
other speakers and other kinds of extraneous factor. Since you seem to
be using Dutch-Afrikaans as a benchmark, I might mention that a native
Dutch speaker of my acquaintance told me not long ago that he could
understand hardly anything of Afrikaans.
If you could show consistency of such judgments across a sample of
speakers, you might have something. Correlation of such a consistent
judgment with a quantitative measure (such as you seem to be assuming
above) would be worth publishing.
Anyhow, if you have been getting these judgments from internet
contacts, rather than trudging through the potato fields of rural
Germany, then presumably your judgments are coming from people who
know the national standard, and are therefore less likely to be good
indicators of direct dialect-to-dialect intelligibility.
> > 2) "80% ... 90 %..."
>
> If you read through Ethnologue, they seem to be splitting these days
> at 90%. And you read all the literature on intelligibility done by SIL
> people, this is indeed what they say. They split at 89%.
I'm aware that SIL people have such tests. What I wanted to know was
whether your figures were derived from such tests, and whether done by
yourself or others.
> > 3) "just by looking"
>
> Not so difficult! Let us take a look here: Hamburgisch, Ollands and
> Oldenburg, 3 Low Saxon lects:
>
> Hamburgs:
>
> http://www.lowlands-l.net/anniversary/hamborgsch.php
>
> Ollands:
>
> http://www.lowlands-l.net/anniversary/ollandsch.php
>
> Oldenburg:
>
> http://www.lowlands-l.net/anniversary/ollenborgsch.php
>
> Quick observation shows us that Hamburgs and Ollands obviously must be
> dialects of one tongue. Yet Oldenburg seems so different that it
> doesn't seem possible for them to converse with the others at 90%+.
>
> Keep in mind that much of science is simply observational, hunches,
> intuition, etc. Sir William Jones famous discovery certainly was.
>
I don't question that hunches, impressions, etc can be right. It's
just that rather than presenting us with a provocative and fruitful
hypothesis (like Jones), you seem to be presenting discoveries -- that
X and Y are "really" different languages, when I'm not sure that
concept is solid enough to make that a discovery. If you say German is
"really" 84 different languages, it could be that you are just using
the terms different/language in an idiosyncratic way.
> > So (1) you talk to large numbers of people about dialect differences?
> > (2) You (or somebody) has applied some standard quantitative test to
> > many pairs of dialects?
>
> Yes, tests have been done and the results are that most German
> "dialects" are in fact separate languages. Swabish, Bavarian, Swiss
> German, Mainfrankisch, Pfalzisch, Luxembourgish, Hessian, Badisch all
> have 40% intelligibility with Standard German.
> Hutterite has 70% intelligibility with Pennsylvania German and 50%
> with St German and Plautdietsch (Mennonite).
And these are on the same SIL tests?
> Problem is that to do this testing you usually need a funding source
> and a team to go out and do it. I'd love to see lots more testing done
> in German. In Chinese, too. Lots of places. Hardly anyone does this
> for some reason, maybe because it's boring. Linguists are usually off
> dabbling in highfalutin theory. Most testing nowadays is being done by
> SIL by Bible translation purposes.
It would be interesting to know why you don't find it "boring".
>
> > (3) Some of it is just impressions?
>
> Some, not much.
>
> A lot also comes from the linguistic literature.
If I was interested enough I might ask for some references to that
literature. I don't doubt you've got them, but I'm not likely to get
around to looking at them any time soon.
Ross Clark
Ruud Harmsen
<benl...@ihug.co.nz>: in sci.lang:
>Statements of this kind are notoriously influenced by attitudes to
>other speakers and other kinds of extraneous factor. Since you seem to
>be using Dutch-Afrikaans as a benchmark, I might mention that a native
>Dutch speaker of my acquaintance told me not long ago that he could
>understand hardly anything of Afrikaans.
It so happens I was watching a dvd about writer W.F. Hermans, which
contains some bilungual interviews. There were three speakers, one of
which probably spoke Dutch with an Afrikaans accent, but the other two
spoke geniune Afrikaans. It wasn't so easy to understand for me, even
with Dutch subtitles.
--
Ruud Harmsen, http://rudhar.com
Bob
> Bob wrote:
> > On Mar 9, 7:58 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> >> On Mar 10, 12:14 pm, Bob <lindsay.rob...@gmail.com> wrote:
> >>> On Mar 9, 1:57 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> >>>> On Mar 10, 5:51 am, Bob <lindsay.rob...@gmail.com> wrote:
> >>>>>http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
> > Exactly, this is how Ethnologue, and others, typically split many
> > languages. You ask speakers. Hey, can you understand those people over
> > there? "Sure." Dialects. "Are you kidding?" Separate languages.
> > Something in between? Much trickier, but often speakers will attempt
> > to clarify via percentages how much they understand.
>
> It's laughably unrealistic to base that on such provably flimsy
> value judgments.
>
> You'd need "standard" (whatever that means) speakers
> with IQ=100, all with the same understanding of what
> "understand those people" mean, and with no prior exposure
> to "those peoples'" language.
Maybe so, but this is typically how it is done by Ethnologue and
others. Linguistics is a science, and if people are wrong about lects
being languages or dialects, sooner or later it all sorts out. For
instance, Ethnologue recently was petitioned by Mesoamerican
linguistics experts to get rid of numerous Maya languages that they
had split as separate languages. So it looks like SIL is going to lump
these together. If you study the literature, "whether or not people
can understand each other" is not really very controversial. And I
have been through all of the intelligibility lit, unlike all of you, I
guess.
State of the art here:
http://www.ethnologue.com/show_work.asp?id=10386
Casad, Eugene H. 1974. Dialect intelligibility testing. Summer
Institute of Linguistics Publications in Linguistics and Related
Fields, 38. Norman: Summer Institute of Linguistics of the University
of Oklahoma. xiv, 201 p.
If all else fails, there are tests. If you review the literature, an
extremely science based approach has been moving along in
intelligibility assessment, with many critiques of validity,
reliability, etc, and endless reviving of hypotheses, conclusions,
etc. To say that there is no scientific way that the science of
linguistics can determine intelligibility is the pinnacle of
obscurantism and Know Nothingism.
This problem has been dealt with in the lit. This is the very thorny
problem of learned bilingualism, and yes, it IS a problem. SIL (and I
know some of the folks who sit on the committee that determines what's
a language and what's not), when they denote "inherent
intelligibility" appear to be referring to "virgin ears". But the
problem of learned bilingualism is very difficult to assess. This is
clearly what we are dealing with with Czech and Slovak above. The
trick here is to find the "virgin ears" since learned bilingualism
makes no meaningful statements about intelligibility.
Say you find a village in Mexico where you have five males and five
females that you decide to test. The males test about 90% on the Lect
2 and the females test 50%. The intelligibility is 70% and you split
to languages, since SIL seems to split at 90%. See the Mexico page
here for evidence that splitting begins at 90%:
http://www.ethnologue.com/show_country.asp?name=MX
>
> > Once they start
> > saying, "We find it really hard to understand those people," you're
> > almost always dealing with 2 languages. At 90%+, people usually say
> > they can understand them well, more or less, etc. When they say things
> > like, "We can understand them partly but not fully," you are often
> > dealing with ~60-70% intelligibility.
>
> That's like conducting an IQ test, where you don't make
> people solve the puzzles and give you objective answers,
> instead just tell you how easy they think each question is
> on the scale of 0 to 100%. :-)))
Well, the best way of all is intelligibility testing, but this is
probably not carried out with most languages, perhaps unfortunately.
So the state of the science is not that good. But if you study
intelligibility testing a lot, you find that when they do the tests,
it pretty much confirms what the people were telling you anyway.
This is what SIL does. Fist you need to ask quite a few folks, and
interview them separately, because group interviews fail due to a
strong personality who convinces others to maybe lie. You must
interview alone.
Can you understand those people? "Sure, no problem." Dialect, move on.
"No way, are you kidding?" Separate languages. Well, sort of. In that
case, Swadesh list. Swadesh list is used to sort the wheat from the
chaff, but Swadesh list is no good for intelligibility testing.
However, at <70%, you've got two languages, time to move on. At >90%
on Swadesh, they assume dialects and move on. 70-89% on Swadesh is the
key. Now you need to move on to intelligibility testing.
Intelligibility testing has been revised endlessly in a rigorous
empirical environment for over 50 years since Morris Swadesh wrote the
first piece in the 1950's.
Common techniques include sentence repetition, etc. You interview a #
of folks and average it all together.
You guys act like this is going to throw up all these false
conclusions about intelligibility, but I haven't heard of one yet. Can
you toss out an example?
Bob
> On Mar 10, 5:47 pm, Bob <lindsay.rob...@gmail.com> wrote:
>
>
>
> > On Mar 9, 7:58 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>
> > > On Mar 10, 12:14 pm, Bob <lindsay.rob...@gmail.com> wrote:
>
> > > > On Mar 9, 1:57 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
>
> > > > > On Mar 10, 5:51 am, Bob <lindsay.rob...@gmail.com> wrote:
>
> > > > > >http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
> > Exactly, this is how Ethnologue, and others, typically split many
> > languages. You ask speakers. Hey, can you understand those people over
> > there? "Sure." Dialects. "Are you kidding?" Separate languages.
> > Something in between? Much trickier, but often speakers will attempt
> > to clarify via percentages how much they understand. Once they start
> > saying, "We find it really hard to understand those people," you're
> > almost always dealing with 2 languages. At 90%+, people usually say
> > they can understand them well, more or less, etc. When they say things
> > like, "We can understand them partly but not fully," you are often
> > dealing with ~60-70% intelligibility.
>
> Statements of this kind are notoriously influenced by attitudes to
> other speakers and other kinds of extraneous factor. Since you seem to
> be using Dutch-Afrikaans as a benchmark, I might mention that a native
> Dutch speaker of my acquaintance told me not long ago that he could
> understand hardly anything of Afrikaans.
> If you could show consistency of such judgments across a sample of
> speakers, you might have something. Correlation of such a consistent
> judgment with a quantitative measure (such as you seem to be assuming
> above) would be worth publishing.
> Anyhow, if you have been getting these judgments from internet
> contacts, rather than trudging through the potato fields of rural
> Germany, then presumably your judgments are coming from people who
> know the national standard, and are therefore less likely to be good
> indicators of direct dialect-to-dialect intelligibility.
No, "not intelligible with St German" doesn't cut it for separate
language. If you're going to do that, you end up with 4,000 separate
languages in Low Saxon alone! That's how many dialects there are in
Low Saxon. They're practically all not intelligible with St German.
That's not enough. We want to know that, in addition to that, German
Dialect A can't understand German Dialect B. You need to know their
names and all the details. For example, there may well be 40 separate
languages in Swiss German, but until someone can tell me that Bern
can't understand Zurich or whatever, we can't do anything, if you use
an evidence-based approach.
The Dutch speaker is clearly an outlier. I read through a # of
description of Dutch speakers hearing Afrikaans, and the average
seemed around ~80% (writing was almost 100%). What you do is talk to a
lot of speakers, then average them or weed out outliers.
>
> > > 2) "80% ... 90 %..."
>
> > If you read through Ethnologue, they seem to be splitting these days
> > at 90%. And you read all the literature on intelligibility done by SIL
> > people, this is indeed what they say. They split at 89%.
>
> I'm aware that SIL people have such tests. What I wanted to know was
> whether your figures were derived from such tests, and whether done by
> yourself or others.
In some cases they are, but in general, SIL are the only folks doing
these tests, and they've only tested some German lects - incidentally,
typical distance between Average German Dialect and St German was
~80%. Testing has been inside Ripaurian and has found that at either
ends of the Ripaurian dialect chain, intelligibility is ~20%. So that
right there makes the notion of one Ripaurian language nuts.
It would be great if more testing would be done. It's good hard
science for intelligibility.
>
> > Keep in mind that much of science is simply observational, hunches,
> > intuition, etc. Sir William Jones famous discovery certainly was.
>
> I don't question that hunches, impressions, etc can be right. It's
> just that rather than presenting us with a provocative and fruitful
> hypothesis (like Jones), you seem to be presenting discoveries -- that
> X and Y are "really" different languages, when I'm not sure that
> concept is solid enough to make that a discovery.
No, not at all. They are really hypotheses. Are you familiar with
"pilot study"? I'm just throwing out some provocative hypotheses, and
hoping for lots of further thinking, study and *criticism* will come
of that. If I'm wrong, fantastic, but someone needs to get the ball
rolling here.
If you say German is
> "really" 84 different languages, it could be that you are just using
> the terms different/language in an idiosyncratic way.
I'm sorry. I'm presenting a very weak hypothesis that German may be 90
languages and hoping for tons of critical response to clear things up.
> > Yes, tests have been done and the results are that most German
> > "dialects" are in fact separate languages. Swabish, Bavarian, Swiss
> > German, Mainfrankisch, Pfalzisch, Luxembourgish, Hessian, Badisch all
> > have 40% intelligibility with Standard German.
> > Hutterite has 70% intelligibility with Pennsylvania German and 50%
> > with St German and Plautdietsch (Mennonite).
>
> And these are on the same SIL tests?
Those are all SIL figures I think.
>
> > Problem is that to do this testing you usually need a funding source
> > and a team to go out and do it. I'd love to see lots more testing done
> > in German. In Chinese, too. Lots of places. Hardly anyone does this
> > for some reason, maybe because it's boring. Linguists are usually off
> > dabbling in highfalutin theory. Most testing nowadays is being done by
> > SIL by Bible translation purposes.
>
> It would be interesting to know why you don't find it "boring".
I'm fascinated by it. Looking around the net, tons of non-linguists
are. I'm not sure why academics often shy away from this stuff.
Probably because an academic writing what I wrote is really sticking
his neck out to get pounded. Anyway, I know these guys, and they are
mostly presenting papers on Optimality Theory or Generative Syntax and
they don't mess around with this stuff.
I know some of these guys, and they don't seem to care if some lect is
a language or a dialect. If you ask them, they just refer to SIL and
say, SIL is calling this a language nowadays and shrug their heads.
Plus when you start splitting all these languages, you run into all
this political BS from nationalists screaming that are making
languages out of "dialects" of the national tongue, or fostering
separatism or whatever. Look around the Net for the wild debate about
whether Scots is a dialect of English or a language. I mean, a lot of
people just go stark raving furious when you say Scots is a separate
language for some reason.
Can you imagine the media firestorm if SIL decided that AAVE was a
separate language from St English?
Speaking of which, why don't we split Geordie off? Probably political
reasons.
> If I was interested enough I might ask for some references to that
> literature. I don't doubt you've got them, but I'm not likely to get
> around to looking at them any time soon.
There should be some links in the piece.
>
> Ross Clark
grammatim
> On Mar 10, 12:46 am, "PaulJK" <paul.kr...@gmail.com> wrote:
>
>
>
>
>
> > Bob wrote:
> > > On Mar 9, 7:58 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> > >> On Mar 10, 12:14 pm, Bob <lindsay.rob...@gmail.com> wrote:
> > >>> On Mar 9, 1:57 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> > >>>> On Mar 10, 5:51 am, Bob <lindsay.rob...@gmail.com> wrote:
> > >>>>>http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
> > > Exactly, this is how Ethnologue, and others, typically split many
> > > languages. You ask speakers. Hey, can you understand those people over
> > > there? "Sure." Dialects. "Are you kidding?" Separate languages.
> > > Something in between? Much trickier, but often speakers will attempt
> > > to clarify via percentages how much they understand.
>
> > It's laughably unrealistic to base that on such provably flimsy
> > value judgments.
>
> > You'd need "standard" (whatever that means) speakers
> > with IQ=100, all with the same understanding of what
> > "understand those people" mean, and with no prior exposure
> > to "those peoples'" language.
>
> Maybe so, but this is typically how it is done by Ethnologue and
> others.
Hos do you know that?
> Linguistics is a science, and if people are wrong about lects
> being languages or dialects, sooner or later it all sorts out. For
_Linguists_ do not care about or address the question of language vs.
dialect. It is purely a political question.
> instance, Ethnologue recently was petitioned by Mesoamerican
> linguistics experts to get rid of numerous Maya languages that they
> had split as separate languages. So it looks like SIL is going to lump
> these together. If you study the literature, "whether or not people
> can understand each other" is not really very controversial. And I
> have been through all of the intelligibility lit, unlike all of you, I
> guess.
Tell us what literature you have studied on questions of mutual
intelligibility. With full bibliographic references so other people
can check.
> State of the art here:
>
> http://www.ethnologue.com/show_work.asp?id=10386
>
> Casad, Eugene H. 1974. Dialect intelligibility testing. Summer
> Institute of Linguistics Publications in Linguistics and Related
> Fields, 38. Norman: Summer Institute of Linguistics of the University
> of Oklahoma. xiv, 201 p.
So nothing has been done in this area in 35 years, despite the immense
progress in variationist linguistics?
> If all else fails, there are tests. If you review the literature, an
> extremely science based approach has been moving along in
> intelligibility assessment, with many critiques of validity,
> reliability, etc, and endless reviving of hypotheses, conclusions,
> etc. To say that there is no scientific way that the science of
> linguistics can determine intelligibility is the pinnacle of
> obscurantism and Know Nothingism.
Ok, where's your literature review? Is it in an Annual Review of
Anthropology somewhere?
Ah. So you now admit that your motivation is religious.
No references, of course.
> This is what SIL does. Fist you need to ask quite a few folks, and
> interview them separately, because group interviews fail due to a
> strong personality who convinces others to maybe lie. You must
> interview alone.
>
> Can you understand those people? "Sure, no problem." Dialect, move on.
> "No way, are you kidding?" Separate languages. Well, sort of. In that
> case, Swadesh list. Swadesh list is used to sort the wheat from the
> chaff, but Swadesh list is no good for intelligibility testing.
> However, at <70%, you've got two languages, time to move on. At >90%
> on Swadesh, they assume dialects and move on. 70-89% on Swadesh is the
> key. Now you need to move on to intelligibility testing.
> Intelligibility testing has been revised endlessly in a rigorous
> empirical environment for over 50 years since Morris Swadesh wrote the
> first piece in the 1950's.
Yet you haven't offered a single reference less than 35 years old.
> Common techniques include sentence repetition, etc. You interview a #
> of folks and average it all together.
>
> You guys act like this is going to throw up all these false
> conclusions about intelligibility, but I haven't heard of one yet. Can
> you toss out an example?-
Linguists don't care about the lines you are so eager to draw.
grammatim
> In some cases they are, but in general, SIL are the only folks doing
> these tests, and they've only tested some German lects - incidentally,
> typical distance between Average German Dialect and St German was
> ~80%. Testing has been inside Ripaurian and has found that at either
> ends of the Ripaurian dialect chain, intelligibility is ~20%. So that
> right there makes the notion of one Ripaurian language nuts.
You can't assert that until you have some non-subjective definition of
your terms.
> I'm fascinated by it. Looking around the net, tons of non-linguists
> are. I'm not sure why academics often shy away from this stuff.
Have you never heard of dialectology, and the great Linguistic Atlas
projects?
> Probably because an academic writing what I wrote is really sticking
> his neck out to get pounded. Anyway, I know these guys, and they are
> mostly presenting papers on Optimality Theory or Generative Syntax and
> they don't mess around with this stuff.
Then you know very, very few linguists.
> I know some of these guys, and they don't seem to care if some lect is
> a language or a dialect. If you ask them, they just refer to SIL and
> say, SIL is calling this a language nowadays and shrug their heads.
Precisely. "Language vs. dialect" is not a question of linguistics.
> Plus when you start splitting all these languages, you run into all
> this political BS from nationalists screaming that are making
> languages out of "dialects" of the national tongue, or fostering
> separatism or whatever. Look around the Net for the wild debate about
> whether Scots is a dialect of English or a language. I mean, a lot of
> people just go stark raving furious when you say Scots is a separate
> language for some reason.
>
> Can you imagine the media firestorm if SIL decided that AAVE was a
> separate language from St English?
Precisely. It's exclusively a political question.
> There should be some links in the piece.
(a) What "piece"?
(b) You don't know? You haven't read it??
noesy_parker
257624...@o36g2000yqh.googlegroups.com:
>
> Precisely. "Language vs. dialect" is not a question of linguistics.
>
It's utterances like that that showed linguistics is not a science.
Bob
> On Mar 10, 7:43 am, Bob <lindsay.rob...@gmail.com> wrote:
>
> > I'm fascinated by it. Looking around the net, tons of non-linguists
> > are. I'm not sure why academics often shy away from this stuff.
>
> Have you never heard of dialectology, and the great Linguistic Atlas
> projects?
Those often don't split languages and dialects much. The Linguistic
Atlas of China is an endless list of "dialects."
>
> > Probably because an academic writing what I wrote is really sticking
> > his neck out to get pounded. Anyway, I know these guys, and they are
> > mostly presenting papers on Optimality Theory or Generative Syntax and
> > they don't mess around with this stuff.
>
> Then you know very, very few linguists.
>
> > I know some of these guys, and they don't seem to care if some lect is
> > a language or a dialect. If you ask them, they just refer to SIL and
> > say, SIL is calling this a language nowadays and shrug their heads.
>
> Precisely. "Language vs. dialect" is not a question of linguistics.
Really? Why then a committee of linguists at SIL determine what is a
language and what is a dialect, if linguistics has zero to do with it.
Why do linguists the world over submit ISO apps to SIL to split off
new languages, or get rid of some languages and rephrase them as
dialects.
Is Lyle Campbell a linguist? I will answer that myself. He is one of
the world's top linguists. Campbell and other Americanists submitted
an ISO app to SIL recently to eliminate some Mayan language and
reclassify them as dialects of some larger Mayan languages. If this
question has nothing to do with linguistics, why did Dr. Campbell do
this. BTW, Campbell seems to be on good terms with the SIL crew. I
assume Dr. Campbell is a religious nut too.
Bob
> On Mar 10, 7:24 am, Bob <lindsay.rob...@gmail.com> wrote:
>
>
>
> > On Mar 10, 12:46 am, "PaulJK" <paul.kr...@gmail.com> wrote:
>
> > > Bob wrote:
> > > > On Mar 9, 7:58 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> > > >> On Mar 10, 12:14 pm, Bob <lindsay.rob...@gmail.com> wrote:
> > > >>> On Mar 9, 1:57 pm, "benli...@ihug.co.nz" <benli...@ihug.co.nz> wrote:
> > > >>>> On Mar 10, 5:51 am, Bob <lindsay.rob...@gmail.com> wrote:
> > > >>>>>http://robertlindsay.wordpress.com/2009/03/08/a-reworking-of-german-l...
>
> Linguists don't care about the lines you are so eager to draw.
If "linguists" don't care about what's a language and what's a
dialect, why are some of the world's top and best linguists always
submitting ISO apps to SIL to create new languages, merge others,
eliminate languages and reclassify them as dialects, etc. Is there
something wrong with these people, since "linguists don't care." Are
they crazy, are they not linguists, is something wrong with them?
grammatim
> On Mar 10, 6:06 am, grammatim <gramma...@verizon.net> wrote:
> > On Mar 10, 7:43 am, Bob <lindsay.rob...@gmail.com> wrote:
>
> > > I'm fascinated by it. Looking around the net, tons of non-linguists
> > > are. I'm not sure why academics often shy away from this stuff.
>
> > Have you never heard of dialectology, and the great Linguistic Atlas
> > projects?
>
> Those often don't split languages and dialects much. The Linguistic
> Atlas of China is an endless list of "dialects."
I don't know what that could mean, but if it is, then it isn't a
linguistic atlas in any recognizable sense. The great linguistic
atlases of the first half of the twentieth century are enormous
collections of data -- gathered at immense expense -- plotted on
incredibly detailed large-scale maps, from which people like you can
gather the evidence (which will show that there are usually no lines
on the ground where you want to draw them).
> > > Probably because an academic writing what I wrote is really sticking
> > > his neck out to get pounded. Anyway, I know these guys, and they are
> > > mostly presenting papers on Optimality Theory or Generative Syntax and
> > > they don't mess around with this stuff.
>
> > Then you know very, very few linguists.
>
> > > I know some of these guys, and they don't seem to care if some lect is
> > > a language or a dialect. If you ask them, they just refer to SIL and
> > > say, SIL is calling this a language nowadays and shrug their heads.
>
> > Precisely. "Language vs. dialect" is not a question of linguistics.
>
> Really? Why then a committee of linguists at SIL determine what is a
> language and what is a dialect, if linguistics has zero to do with it.
I've no idea. SIL is a religious missionary organization whose aim is
to translate the Bible into as many languages as possible, believing
that this will hasten the Second Coming. It is very much to their
advantage to multiply the number of "languages." The SIL listing is
universally recognized as the "extreme splitter" case, while area
specialists have to sew back together all the different named entities
they have registered.
> Why do linguists the world over submit ISO apps to SIL to split off
> new languages, or get rid of some languages and rephrase them as
> dialects.
Because two or three years ago SIL became the organization assigned to
care for the ISO listing, over the objection of a number of people who
were opposed to the christianization of the list of the world's
languages.
> Is Lyle Campbell a linguist? I will answer that myself. He is one of
> the world's top linguists. Campbell and other Americanists submitted
> an ISO app to SIL recently to eliminate some Mayan language and
> reclassify them as dialects of some larger Mayan languages. If this
> question has nothing to do with linguistics, why did Dr. Campbell do
> this. BTW, Campbell seems to be on good terms with the SIL crew. I
> assume Dr. Campbell is a religious nut too.
I've never met him. I know that he is a rabid opponent of Greenbergian
megalocomparison (as Jim Matisoff called it), and as a specialist in
Maya linguistics, is probably familiar enough with all 20-odd living
varieties, and their speakers -- who the question is ultimately up to
-- to say what counts as a language and what doesn't, in the opinions
of the speakers.
grammatim
They're not doing linguistics. The ISO is not a linguistic
organization. They're doing politics.
There are no linguistic criteria for the distinction "language" vs.
"dialect."
António Marques
> You're talking about these things as though they're mutually
> exclusive and as though the lines you're drawing are particularly
> meaningful.
*That's* the key issue with Bob's (and other folks') obsessions. They
don't grasp the central concepts of meaningfulness and usefulness.
84? Heck. Make it 480 separate languages. What does any speaker of one
of them care?
--
António Marques
Harlan Messinger
I offer a conjecture (sheer speculation, not something for which I would
argue, but something that I think others have suggested here) that
Ethnologue has a vested interest in teasing out as many languages as
possible: to drive an increase in the number of versions of the Bible
they can give the appearance of needing to publish. The more versions
they can justify, the greater they can impress on contributors and
potential contributors the importance of their mission.
Also, imagine giving a person in Lower Slobovia his very own Bible
translated into Upper West-Central Lower Slobovian. This show of special
attention might give the Upper West-Central Lower Slobovian speaker a
warm, fuzzy feeling that could facilitate proselytization.
António Marques
> ... this is typically how it is done by Ethnologue and others.
Yes, and it's one of the reasons people (except for Wikipedia) don't
rely on Ethnologue's classifications.
> Linguistics is a science, and if people are wrong about lects being
> languages or dialects, sooner or later it all sorts out. For
> instance, Ethnologue recently was petitioned by Mesoamerican
> linguistics experts to get rid of numerous Maya languages that they
> had split as separate languages. So it looks like SIL is going to
> lump these together.
And the picture emerges: whenever someone looks at the data with seeing
eyes, Ethnologue's original approach is found out to be useless. Of
course, some parts of Ethnologue have been vetted; others have not. For
instance, Ethnologue has reported 'galician' (15k), 'asturian' (25k) and
'Miranda do Douro' (15k) for Portugal like since ever. 'Galician' simply
just doesn't exist here, nor has it ever (except for possible
immigrants, of course; but they've even drawn a completely fictitious
map), while the 'asturian' consists of, precisely, nothing else than
'mirandese' (not 'Miranda do Douro' , which is the portuguese name of
the county's capital). Now, when will this be corrected? Never, of
course. For occitan, besides the stupidity of separating it into 6
languages (when actual speakers don't do any such thing and merrily
communicate with each other), they put it inside 'ibero-romance' - for
crying out loud, they have no idea what 'ibero-romance' even is. Their
sources are scant, outdated, poorly chosen, not representative, and they
lack the baggage to evaluate them. And who, in God's name, has the
patience to correct them? Let them be.
The dialect/language distinction was never scientific and to try to make
it seem so is but an exercise on vanity.
--
António Marques
António Marques
> Plus when you start splitting all these languages, you run into all
> this political BS from nationalists screaming that are making
> languages out of "dialects" of the national tongue, or fostering
> separatism or whatever.
It's the opposite, Bob. You run into all the political BS that you're
downgrading their 'language' to a dialect.
I've never seen any english person worried that scots is considered a
separate language, I've seen plenty of scots take offense to the idea
that Scots is 'just a dialect'. (And reasonably so, because the *useful*
thing to do with Scots is precisely to cultivate it as separate from
english.)
--
António Marques
António Marques
Why? Is astronomy any less of a science just because it was only
recently that a committee chose a definition of 'planet', which moreover
is of extermely limited impact?
--
António Marques
noesy_parker
$rp8$2...@nntp.motzarella.org:
The astronomers aren't saying whether something is a planet or not is a
question for astronomy, in fact they take the trouble to make a decision
on something of no great import. Very unlike linguists who appeared to
have turn their back on a problem related to their discipline.
I haven't heard a single valid reasons why language/dialect demarcation
should not be a question for linguistics. Some of reasons given are
plainly assinine, like the point about familiarity and exposure in
relation to Czech/Slovak. This is like, DUH! If linguists can't even
deal with something like that, then there is no hope for it as a
scientific discipline.
Richard Herring
[sorry, soc.culture.* stripped by my server]
In message <71nimiF...@mid.individual.net>, Harlan Messinger
<hmessinger...@comcast.net> writes
>António Marques wrote:
>> Harlan Messinger wrote:
>>> You're talking about these things as though they're mutually
>>> exclusive and as though the lines you're drawing are particularly
>>> meaningful.
>> *That's* the key issue with Bob's (and other folks') obsessions.
>>They
>> don't grasp the central concepts of meaningfulness and usefulness.
>> 84? Heck. Make it 480 separate languages. What does any speaker of
>>one of them care?
>
>I offer a conjecture (sheer speculation, not something for which I
>would argue, but something that I think others have suggested here)
>that Ethnologue has a vested interest in teasing out as many languages
>as possible: to drive an increase in the number of versions of the
>Bible they can give the appearance of needing to publish. The more
>versions they can justify, the greater they can impress on contributors
>and potential contributors the importance of their mission.
And the primary unstated purpose of all organizations is to perpetuate
themselves. Yet logically, the _fewer_ the languages, the sooner their
mammoth task would be complete. Anyone would think they don't actually
deep-down believe in their mission...
--
Richard Herring
Harlan Messinger
Chemists who specialize in pigments have no interest in and reject
responsibility for dividing the Impressionists up into sub-genres based
on differences among the characteristics of their respective painting
styles. If chemists can't even deal with something like that, then there
is no hope for chemistry as a scientific discipline.
António Marques
> =?UTF-8?B?QW50w7NuaW8gTWFycXVlcw==?=<m....@sapo.pt> wrote in news:gp6543
> $rp8$2...@nntp.motzarella.org:
>
>> noesy_parker wrote:
>>> grammatim<gram...@verizon.net> wrote in
>>> news:a852fd21-1f3f-4e03-8167-
>>> 257624...@o36g2000yqh.googlegroups.com:
>>>
>>>> Precisely. "Language vs. dialect" is not a question of
>>>> linguistics.
>>> It's utterances like that that showed linguistics is not a science.
>> Why? Is astronomy any less of a science just because it was only
>> recently that a committee chose a definition of 'planet', which
> moreover
>> is of extermely limited impact?
>
> The astronomers aren't saying whether something is a planet or not is a
> question for astronomy, in fact they take the trouble to make a decision
> on something of no great import. Very unlike linguists who appeared to
> have turn their back on a problem related to their discipline.
So you don't take exception to the fact that a question is decided or
important or not, but to people saying the question is out of the scope
of their field of study?
It's not like Mercedes spends a whole lot of time anguishing over
whether it is a car or truck manufacturer, or what the precise
scientific distinction is between an SUV and an AUV.
> I haven't heard a single valid reasons why language/dialect demarcation
> should not be a question for linguistics.
Language/dialect demarcation is not a question for linguistics to
*decide* for the exact same reason that government
reelection/overthrowal is not a question for social science to decide.
It's not a matter of arguing whether it should be decided or not by
lingusitics, it's the fact that it *is* decided out of science
altogether. By those, err, 'people', I think it is, who use them (and
that includes people who study languages which nobody speaks).
Now, linguistics may study why and how such a demarcation appears or
not, but that's a completely different issue.
--
António Marques