The theory of Mr. Carrasquer

56 views
Skip to first unread message

Andrew Usher

unread,
Oct 4, 2004, 3:02:41 AM10/4/04
to
I have become interested in the origins of Indoeuropean language, and
as I looked at it I thought that linguistic dogmas were preventing the
truth from being researched properly. I began to create my own theory
(without having studied enough history, I don't think) but abandoned
it some time ago when I discoved the remarkable theory expounded by
Miguel Carrasquer in this group from 1996-98.

There seems to be no source to which I could go to get his complete
argument, of which only pieces were presented here. I will thus add my
own speculation.

It would seem to me that before the invention of agriculture (and
civilsation in general) no one would need to learn any foreign
language, and thus great linguistic diversity could arise. This
definitely allow Carrasquer's 6 proto-families (~8000 BC) around the
Fertile Crescent, and that's only those surviving down to historical
time. I say then we will not be able to trace them any farther back,
though certainly they did have a unity once.

I have seen that there is an amateur linguist, Glenn Gordon, who also
believes Tyrrhenian to be closest to IE, though he has a quite
different conception of the origin of Indo-Tyrrhenian - for me, this
lends a bit more support. The fact, also, that it can be reconciled to
all the archaeological evidence is decisive, as well. A premiss of
this hypothesis is that IT was the only family to expand into Europe
in prehistoric times, and as such we should find no evidence of
Semitic language being ever spoken on the mainland - is this the
general consensus?

One of the 'myths' is that which opposes any detailed reconstruction
of a proto-language. Recent works by linguists I have inspected are
alarmingly vague about PIE; many refuse to even say that we can
reconstruct the actual sounds. I do think that linguistics is a
science; that scientific endeavor need not be directed toward anything
practical or of service to other fields of study; and that detailed
reconstruction of proto-languages therefore may well be pursued, and
that a marker of sufficiently good evidence is the ability to compose
text in the proto-language.

Andrew Usher

John Lawler

unread,
Oct 4, 2004, 10:11:44 AM10/4/04
to
Andrew Usher <k_over...@yahoo.com> writes:

>I have become interested in the origins of Indoeuropean language, and
>as I looked at it I thought that linguistic dogmas were preventing the
>truth from being researched properly. I began to create my own theory
>(without having studied enough history, I don't think) but abandoned

^^^^^^^^^^^^^
(parenthetically, a fascinating construction here)

>it some time ago when I discoved the remarkable theory expounded by
>Miguel Carrasquer in this group from 1996-98.

>There seems to be no source to which I could go to get his complete
>argument, of which only pieces were presented here. I will thus add my
>own speculation.

>It would seem to me that before the invention of agriculture (and
>civilsation in general) no one would need to learn any foreign
>language, and thus great linguistic diversity could arise.

Whoa! Stop here. This assumption is simply false. Why in the world would
one assume that? The norm in human culture is for people to know and speak
many languages. That's the way it is almost everywhere in the world,
*especially* in hunting and gathering cultures. It's one of the results of
exogamy, for one thing.

Certainly there is (and was) great linguistic diversity, but there are
always lots of reasons why people should know and speak several languages.
And they do, everywhere except in the Anglophone world. Possibly you may
need to do a little more reading in basic linguistics.

Oh, and if you're interested in PIE, there are two books you need to get:

o 'A Dictionary of Selected Synonyms in the Principal Indo-European
Languages', by Carl Darling Buck. University of Chicago Press.

o 'American Heritage Dictionary of Proto-Indo-European Roots' (2nd ed.),
by Calvert Watkins. Houghton Mifflin.

Both are paperback, both are authoritative, and both are fascinating.
When you're familiar with their contents, and how they were arrived at,
c'mon back and we'll talk about theories.

-John Lawler http://www.umich.edu/~jlawler U Michigan Linguistics Dept
-----------------------------------------------------------------------
"Because in our brief lives, we catch so little of the vastness of
history, we tend too much to think of language as being solid as a
dictionary, with granite-like permanence, rather than as the rampant
restless sea of metaphor that it is." -- Julian Jaynes

Miguel Carrasquer

unread,
Oct 4, 2004, 4:21:11 PM10/4/04
to
On 4 Oct 2004 00:02:41 -0700, k_over...@yahoo.com (Andrew
Usher) wrote:

>I have become interested in the origins of Indoeuropean language, and
>as I looked at it I thought that linguistic dogmas were preventing the
>truth from being researched properly. I began to create my own theory
>(without having studied enough history, I don't think) but abandoned
>it some time ago when I discoved the remarkable theory expounded by
>Miguel Carrasquer in this group from 1996-98.

Refresh my mind, what remarkable theory was that?

I believe now as I did then that the geographical location
of the Proto-Indo-European speakers was in the Early
Neolithic Balkans, and that the main body of Indo-European
speakers ca 5500 started to expand throughout the Northern
European lowlands in what is known to archaeologists as the
LBK (LinearBandKeramik) or Linear Ware culture (an earlier
name is "Danubian culture"). This theory is not uniquely
mine: it was advocated by several archaeologists and
historians before (e.g. I basically got the idea from Colin
McEvedy's Penguin Atlas of Ancient History). It got a lot
of attention in Colin Renfrew's version, which was
unfortunate in a way, since Renfrew managed to get almost
all the linguistic arguments wrong.

>There seems to be no source to which I could go to get his complete
>argument, of which only pieces were presented here. I will thus add my
>own speculation.
>
>It would seem to me that before the invention of agriculture (and
>civilsation in general) no one would need to learn any foreign
>language, and thus great linguistic diversity could arise.

On the contrary: before agriculture _everybody_ had to learn
one or more foreign languages, to communicate with
neighbouring tribes, all of which spoke a different
language. It was only with the advent of agriculture that
population densities rose to such levels that most people
would mostly meet other people speaking the same language in
their lifetime.


=======================
Miguel Carrasquer Vidal
m...@wxs.nl

Jacques Guy

unread,
Oct 5, 2004, 11:55:23 AM10/5/04
to
Miguel Carrasquer wrote:
> It was only with the advent of agriculture that
> population densities rose to such levels that most people
> would mostly meet other people speaking the same language in
> their lifetime.

Not everywhere. Viz Papua-New Guinea and Vanuatu, the former
with some 500 to 700 different languages, the latter with
100 or so (working out at 1000 spearkers per language).
In order never to hear a different language you'd have
to stay within a few miles of your village all your life,
and "a few" can be as little as "one".

mb

unread,
Oct 4, 2004, 11:16:44 PM10/4/04
to
Jacques Guy <jg...@alphalink.com.au> wrote in message news:<4162C3...@alphalink.com.au>...

That's strange. No multilinguals inside the villages? No imported brides or grooms?

Jacques Guy

unread,
Oct 5, 2004, 4:34:23 PM10/5/04
to
mb wrote:

> > In order never to hear a different language you'd have
> > to stay within a few miles of your village all your life,
> > and "a few" can be as little as "one".

> That's strange. No multilinguals inside the villages? No imported brides or grooms?

Look what I wrote: "in order to... you would have..."

As for exogamy, the rules vary from island to island,
village to village. If you were born in Vao, you ALWAYS
marry someone from Wala (an island just a mile away),
and vice versa. The two languages are very different,
and mutually unintelligible.

But marrying is (was) not the only factor. You had
trading. You traded pigs, shell ornaments, woven mats,
songs. Yes, they had invented copyright laws. If you
heard a song and you liked it enough, you would pay
for the right to sing it. It was on a village basis:
village A would buy the copyright from village B,
author of the song. The copyright remained with the
"author-village", in other words, A could not sell it
on. A typical price was a 7-year old tusker (a pig whose
tusks made a full circle), or 100 pounds. Not cheap.

Well, that's what my informant told me.

Andrew Usher

unread,
Oct 4, 2004, 11:58:35 PM10/4/04
to
Miguel Carrasquer <m...@wxs.nl> wrote in message news:<89b3m097ccems59m5...@4ax.com>...

> >I have become interested in the origins of Indoeuropean language, and
> >as I looked at it I thought that linguistic dogmas were preventing the
> >truth from being researched properly. I began to create my own theory
> >(without having studied enough history, I don't think) but abandoned
> >it some time ago when I discoved the remarkable theory expounded by
> >Miguel Carrasquer in this group from 1996-98.
>
> Refresh my mind, what remarkable theory was that?

In messages such as 'Re: Invention of Language', Nov 5, 1996. There
were many others in 1996-98, you may search likely terms in Google, if
you need to refresh your memory.

Basically, the two conjectures I especially noticed were:

1. That Etruscan is part of the IE family.

In the past, I have always read total denials that Etruscan is related
to IE and even suggestions (contrary to the archaeology) that they
were native to Italy.
When I saw the comparison tables you made, I was conviced that
Tyrrhenian was a highly divergent branch of the IE family.

2. That quite a few of the world's language families originated in the
Near East and spread because of the start of farming.

This includes the six families mentioned in the post above referenced.

> I believe now as I did then that the geographical location
> of the Proto-Indo-European speakers was in the Early
> Neolithic Balkans, and that the main body of Indo-European
> speakers ca 5500 started to expand throughout the Northern
> European lowlands in what is known to archaeologists as the
> LBK (LinearBandKeramik) or Linear Ware culture (an earlier
> name is "Danubian culture"). This theory is not uniquely
> mine: it was advocated by several archaeologists and
> historians before (e.g. I basically got the idea from Colin
> McEvedy's Penguin Atlas of Ancient History). It got a lot
> of attention in Colin Renfrew's version, which was
> unfortunate in a way, since Renfrew managed to get almost
> all the linguistic arguments wrong.

Yes. May I find online a decent summary of Renfrew's arguments? My
university doesn't have the book (!).

> >There seems to be no source to which I could go to get his complete
> >argument, of which only pieces were presented here. I will thus add my
> >own speculation.
> >
> >It would seem to me that before the invention of agriculture (and
> >civilsation in general) no one would need to learn any foreign
> >language, and thus great linguistic diversity could arise.
>
> On the contrary: before agriculture _everybody_ had to learn
> one or more foreign languages, to communicate with
> neighbouring tribes, all of which spoke a different
> language. It was only with the advent of agriculture that
> population densities rose to such levels that most people
> would mostly meet other people speaking the same language in
> their lifetime.

I must question your assumptions here.

What I had in mind was fairly large areas containing tribes speaking
mutually intelligible dialects. They would certainly meet foreigners
but would not have any reason to learn their language; indeed,
probably the only significant interaction between them was war. If
there were not such areas, and every tribe spoke a different language,
how do you explain those six ancient families; as pidgins? I doubt it.

In all of history until the 20th century most people were monolingual
and that is my default assumption for prehistoric times, as well.

As well, I would think that any nomadic tribe would meet several
hundred others in a human lifetime; do you think any speaker learned
the language of every one? Absurd.

By the way, are you in fact a professional linguist?

Andrew Usher

Andrew Usher

unread,
Oct 5, 2004, 12:00:21 AM10/5/04
to
Jacques Guy <jg...@alphalink.com.au> wrote in message news:<4162C3...@alphalink.com.au>...

> > It was only with the advent of agriculture that

Are you sure about those figures? I am very suspicious that many of
those 'languages' are mutually intelligible dialects. Anthropologists
will naturally be biased toward recognising as many as possible, and
every tribe will insist on the uniqueness of 'their' language.

Andrew Usher

Jacques Guy

unread,
Oct 5, 2004, 5:27:51 PM10/5/04
to
Andrew Usher wrote:

> Are you sure about those figures? I am very suspicious that many of
> those 'languages' are mutually intelligible dialects.

Within an area where I have done intensive fieldwork, a
five-mile stretch on the eastern coast of Espiritu Santo
you have three mutually unintelligible "dialects" if it
pleases to call them dialects. I'll give just the word
for "four" as an example:

1. /Bati/ (B = bilabial fricative)
2. /Tar/ (T = dental fricative, r = trilled r)
3. /iED/ (D = dental fricative)

All three are closely related, yet totally mutually
unintelligible.

For Papua-New Guinea I have to resort to quoting my
late colleague Donald Laycock. Where travel is easy,
every village speaks a dialect unintelligible to its
neighbour. Where travel is difficult, huge areas speak
the same or similar dialects. It seems that where contact
was common each village consciously made its communalect
as different as possible from the next.

I have heard of apparently similar phenomena in
France. The patois of Sarthe is mutually intelligible
with that of Yonne (a long way away: Sarthe is in
Normandy, Yonne in Burgundy), but the geographically
intermediate ones are not.

John Atkinson

unread,
Oct 5, 2004, 1:10:24 AM10/5/04
to

"Andrew Usher" <k_over...@yahoo.com> wrote ...

> Jacques Guy <jg...@alphalink.com.au> wrote...


>
> > > It was only with the advent of agriculture that
> > > population densities rose to such levels that most people
> > > would mostly meet other people speaking the same language in
> > > their lifetime.
> >
> > Not everywhere. Viz Papua-New Guinea and Vanuatu, the former
> > with some 500 to 700 different languages, the latter with
> > 100 or so (working out at 1000 spearkers per language).
> > In order never to hear a different language you'd have
> > to stay within a few miles of your village all your life,
> > and "a few" can be as little as "one".
>
> Are you sure about those figures?

No, because he's wrong. There are around 700 "Papuan" languages, probably a
little more, in New Guinea, but there's also lots of Austronesian languages
there. The Ethnologue lists 823 living languages for Papua-New Guinea and
263 for West Papua. No doubt there's some overlap, but if you said a
thousand total you wouldn't be far wrong.

> I am very suspicious that many of
> those 'languages' are mutually intelligible dialects. Anthropologists
> will naturally be biased toward recognising as many as possible, and
> every tribe will insist on the uniqueness of 'their' language.

Like in Australia, where there are (were, rather) around 250 "languages" by
linguistic criteria, as opposed to seven hundred or more "political"
languages, by which different tribes identified themselves.

"Anthropologists" may, or may not, be biased like that, I wouldn't know.
Linguists, however, don't pay too much attention to them in doing their
counts.

Foley divides the Papuan languages into "upwards of sixty distinct language
families". That's sixty families that, like Indo-European, haven't be
proved to be related to any other using the comparative method. In fact,
though, many of these families are not much more divergent than the Romance
family. Most people expect the number of "unrelated" families to end up
considerably less than sixty once more work has been done. Various more or
less sensible hypotheses exist, from the ridiculous "Indo-Pacific" that
contains all of them and more, through intermediate entities like the
Trans-New-Guinea Phylum, to groupings like Proto-Highlands, for which there
is quite good evidence, more suggestive than probative at this stage
however.

John.


Ruud Harmsen

unread,
Oct 5, 2004, 4:12:15 AM10/5/04
to
>> Jacques Guy <jg...@alphalink.com.au> wrote...

>> > Not everywhere. Viz Papua-New Guinea and Vanuatu, the former
>> > with some 500 to 700 different languages, the latter with
>> > 100 or so (working out at 1000 spearkers per language).

Tue, 05 Oct 2004 05:10:24 GMT: "John Atkinson" <john...@bigpond.com>:
in sci.lang:


>No, because he's wrong. There are around 700 "Papuan" languages, probably a
>little more, in New Guinea, but there's also lots of Austronesian languages
>there. The Ethnologue lists 823 living languages for Papua-New Guinea and
>263 for West Papua. No doubt there's some overlap, but if you said a
>thousand total you wouldn't be far wrong.

He didn't say "a thousand languages", but "1000 speakers per
language".



>Foley divides the Papuan languages into "upwards of sixty distinct language
>families". That's sixty families that, like Indo-European, haven't be
>proved to be related to any other using the comparative method. In fact,
>though, many of these families are not much more divergent than the Romance
>family.

The Romance family as divergent from what?
Or do you mean the lanuage members of the Romance family? If the
Papuan language families are that close, how can it be they haven't
been proven to be related? There seems to be a contradiction here.

--
Ruud Harmsen - http://rudhar.com

Miguel Carrasquer

unread,
Oct 5, 2004, 5:55:11 AM10/5/04
to
On 4 Oct 2004 20:58:35 -0700, k_over...@yahoo.com (Andrew
Usher) wrote:

>Miguel Carrasquer <m...@wxs.nl> wrote in message news:<89b3m097ccems59m5...@4ax.com>...
>
>> >I have become interested in the origins of Indoeuropean language, and
>> >as I looked at it I thought that linguistic dogmas were preventing the
>> >truth from being researched properly. I began to create my own theory
>> >(without having studied enough history, I don't think) but abandoned
>> >it some time ago when I discoved the remarkable theory expounded by
>> >Miguel Carrasquer in this group from 1996-98.
>>
>> Refresh my mind, what remarkable theory was that?
>
>In messages such as 'Re: Invention of Language', Nov 5, 1996. There
>were many others in 1996-98, you may search likely terms in Google, if
>you need to refresh your memory.
>
>Basically, the two conjectures I especially noticed were:
>
>1. That Etruscan is part of the IE family.
>
>In the past, I have always read total denials that Etruscan is related
>to IE and even suggestions (contrary to the archaeology) that they
>were native to Italy.
>When I saw the comparison tables you made, I was conviced that
>Tyrrhenian was a highly divergent branch of the IE family.

I prefer to see Etruscan (and its relatives Lemnian and
perhaps Rhaetic) as descendants of a sister language of PIE,
the common ancestor of which I'd call "Indo-Tyrrhaenian".

>2. That quite a few of the world's language families originated in the
>Near East and spread because of the start of farming.

Farming is one mechanism for language spread. The growth and
expansion of some language families began with farming:
Semitic, Indo-European, perhaps Dravidian. The phenomenon
is not limited to the Near East: the spread of Chinese and
Mayan are examples elsewhere.

>> I believe now as I did then that the geographical location
>> of the Proto-Indo-European speakers was in the Early
>> Neolithic Balkans, and that the main body of Indo-European
>> speakers ca 5500 started to expand throughout the Northern
>> European lowlands in what is known to archaeologists as the
>> LBK (LinearBandKeramik) or Linear Ware culture (an earlier
>> name is "Danubian culture"). This theory is not uniquely
>> mine: it was advocated by several archaeologists and
>> historians before (e.g. I basically got the idea from Colin
>> McEvedy's Penguin Atlas of Ancient History). It got a lot
>> of attention in Colin Renfrew's version, which was
>> unfortunate in a way, since Renfrew managed to get almost
>> all the linguistic arguments wrong.
>
>Yes. May I find online a decent summary of Renfrew's arguments? My
>university doesn't have the book (!).

You can get it from amazon used $10, new $25.

>> >There seems to be no source to which I could go to get his complete
>> >argument, of which only pieces were presented here. I will thus add my
>> >own speculation.
>> >
>> >It would seem to me that before the invention of agriculture (and
>> >civilsation in general) no one would need to learn any foreign
>> >language, and thus great linguistic diversity could arise.
>>
>> On the contrary: before agriculture _everybody_ had to learn
>> one or more foreign languages, to communicate with
>> neighbouring tribes, all of which spoke a different
>> language. It was only with the advent of agriculture that
>> population densities rose to such levels that most people
>> would mostly meet other people speaking the same language in
>> their lifetime.
>
>I must question your assumptions here.
>
>What I had in mind was fairly large areas containing tribes speaking
>mutually intelligible dialects. They would certainly meet foreigners
>but would not have any reason to learn their language; indeed,
>probably the only significant interaction between them was war.

Marriage, travel, exchange/trade...

>If there were not such areas, and every tribe spoke a different language,
>how do you explain those six ancient families; as pidgins? I doubt it.

I don't follow. Before the rise of farming, many languages
must have been spoken in the area around the ancient Near
East, just like anywhere else. When farming was "invented",
it was taken up by some of the tribes, and their languages
survived, at least initially (Sumerian, Hurrian, etc.), and
perhaps not by others, who were quickly outnumbered by the
farmers, so that we don't know what languages they spoke.

>In all of history until the 20th century most people were monolingual
>and that is my default assumption for prehistoric times, as well.

That is wrong. Most people in hunter/gatherer societies are
multilingual.

>As well, I would think that any nomadic tribe would meet several
>hundred others in a human lifetime; do you think any speaker learned
>the language of every one? Absurd.

One learnt the language of one's own tribe, the language of
the other moiety of one's own tribe, where applicable, and
one learnt to make oneself understood in the languages of
the most important nearby tribes, especially the ones that
were a source for spouses, and/or in the local lingua
franca, where applicable.

>By the way, are you in fact a professional linguist?

I have a degree in Russian and Slavic literature and
linguistics. I'm not currently being payed for being a
linguist.

Miguel Carrasquer

unread,
Oct 5, 2004, 5:59:26 AM10/5/04
to

There is apparently a difference between the demographic and
linguistic effects of cereal-based temperate agriculture and
tropical agriculture/horticulture.

Language spread over vast areas can also occur _without_
agriculture, as can be seen in Australia (assuming the
Pama-Nyungan family is for real).

benlizross

unread,
Oct 5, 2004, 6:55:23 AM10/5/04
to

I don't find anthropologists to be biased one way or the other.
The figure of 100 for Vanuatu was based on the first comprehensive
language survey of the country, carried out in the 1970s (Tryon 1976).
A more recent survey (Lynch & Crowley 2001) adopts more of a "lumping"
strategy, and gets the figure down to somewhere in the 80s. You won't
get much lower than that. The island of Malakula, quite a bit smaller
than Mallorca, has more than 20 languages. Really different languages.
And multilingualism was the norm, long before the 20th century, for
God's sake.
On the other hand, these people did have agriculture.

Ross Clark

Peter T. Daniels

unread,
Oct 5, 2004, 8:10:31 AM10/5/04
to
Ruud Harmsen wrote:
>
> >> Jacques Guy <jg...@alphalink.com.au> wrote...
> >> > Not everywhere. Viz Papua-New Guinea and Vanuatu, the former
> >> > with some 500 to 700 different languages, the latter with
> >> > 100 or so (working out at 1000 spearkers per language).
>
> Tue, 05 Oct 2004 05:10:24 GMT: "John Atkinson" <john...@bigpond.com>:
> in sci.lang:
> >No, because he's wrong. There are around 700 "Papuan" languages, probably a
> >little more, in New Guinea, but there's also lots of Austronesian languages
> >there. The Ethnologue lists 823 living languages for Papua-New Guinea and
> >263 for West Papua. No doubt there's some overlap, but if you said a
> >thousand total you wouldn't be far wrong.
>
> He didn't say "a thousand languages", but "1000 speakers per
> language".

Right. He said 500-700, which is a 50% to 100% undercount.

> >Foley divides the Papuan languages into "upwards of sixty distinct language
> >families". That's sixty families that, like Indo-European, haven't be
> >proved to be related to any other using the comparative method. In fact,
> >though, many of these families are not much more divergent than the Romance
> >family.
>
> The Romance family as divergent from what?
> Or do you mean the lanuage members of the Romance family? If the
> Papuan language families are that close, how can it be they haven't
> been proven to be related? There seems to be a contradiction here.

He said that each of the sixty-odd PNG families exhibits about as much
diversity as the Romance family -- i.e., not much. But relations between
the families remain to be demonstrated in detail.
--
Peter T. Daniels gram...@att.net

Ruud Harmsen

unread,
Oct 5, 2004, 12:20:14 PM10/5/04
to
Tue, 05 Oct 2004 12:10:31 GMT: "Peter T. Daniels"
<gram...@worldnet.att.net>: in sci.lang:

>> The Romance family as divergent from what?
>> Or do you mean the lanuage members of the Romance family? If the
>> Papuan language families are that close, how can it be they haven't
>> been proven to be related? There seems to be a contradiction here.
>
>He said that each of the sixty-odd PNG families exhibits about as much
>diversity as the Romance family -- i.e., not much. But relations between
>the families remain to be demonstrated in detail.

OK, I misread the expression. Clear now. Thanks.

John Atkinson

unread,
Oct 6, 2004, 6:03:00 AM10/6/04
to

"Jacques Guy" <jg...@alphalink.com.au> wrote...

> Andrew Usher wrote:
>
> > Are you sure about those figures? I am very suspicious that many of
> > those 'languages' are mutually intelligible dialects.
>
> Within an area where I have done intensive fieldwork, a
> five-mile stretch on the eastern coast of Espiritu Santo
> you have three mutually unintelligible "dialects" if it
> pleases to call them dialects. I'll give just the word
> for "four" as an example:
>
> 1. /Bati/ (B = bilabial fricative)
> 2. /Tar/ (T = dental fricative, r = trilled r)
> 3. /iED/ (D = dental fricative)
>
> All three are closely related, yet totally mutually
> unintelligible.
>
> For Papua-New Guinea I have to resort to quoting my
> late colleague Donald Laycock. Where travel is easy,
> every village speaks a dialect unintelligible to its
> neighbour. Where travel is difficult, huge areas speak
> the same or similar dialects. It seems that where contact
> was common each village consciously made its communalect
> as different as possible from the next.

Another way of looking at it, more historical. The "small" languages tend
to be in the low-lying parts of the country. The "large" languages are in
the highlands. Following the last ice age, the sea level rose and drowned
much of the low-lying country (the Sepik valley, for example, now one of the
most linguistically diverse areas, was a big bay.) Most people had to
concentrate in the highlands. Ten thousand years, two things happened:
agriculture was invented, and the lowlands filled up due to erosion. The
crops they had then weren't all that suitable to the highland climate, so
most people came down hill, multiplied, and spread all over. Their
languages did likewise. This is the source of most of the current language
diversity. The highland population remained low, so there weren't that many
languages up there. A few hundred years ago, the sweet potato was
introduced. The highland population exploded, so much so that a few dozen
languages came to be spoken by large numbers of people -- today, Chimbu,
Enga, Western Dani each have more than 100 000, compared with a thousand or
two for a typical lowland language.

In Eurasia, mountains are usually areas of linguistic diversity compared
with lowlands. In New Guinea it's the reverse. Somewhat similarly in South
America, by the way -- possibly (?) due to a similar reason, the population
explosion in the highlands associated with potato cultivation over the last
couple of thousand years.

John.


Douglas G. Kilday

unread,
Oct 6, 2004, 6:09:26 AM10/6/04
to

"Andrew Usher" <k_over...@yahoo.com> wrote ...
>
> [...]

>
> One of the 'myths' is that which opposes any detailed reconstruction
> of a proto-language. Recent works by linguists I have inspected are
> alarmingly vague about PIE; many refuse to even say that we can
> reconstruct the actual sounds. I do think that linguistics is a
> science; that scientific endeavor need not be directed toward anything
> practical or of service to other fields of study; and that detailed
> reconstruction of proto-languages therefore may well be pursued, and
> that a marker of sufficiently good evidence is the ability to compose
> text in the proto-language.

All right, smart guy, let's see you reconstruct archaic Latin text on the
exclusive basis of modern Romance vernacular texts. Let's see you
reconstruct the ablative suffix <-d>, the middle imperative suffix <-mino>,
and subjunctives like <siem>, <edim>, <duim>. Let's see you reconstruct
archaic Latin syntax in its entirety. Let's see you reconstruct the
Saturnian metre. Let's see you pull Latin lexemes out of your magician's
hat which have no reflexes in Romance. In fact, let's see you compose a
realistic archaic Latin epitaph in Saturnians, using only evidence derived
from modern Romance vernacular texts.

Once you have convinced me that you can do these things, I might be willing
to listen to your droning about the feasibility of pursuing detailed
reconstruction of protolanguages.

mb

unread,
Oct 7, 2004, 1:33:41 AM10/7/04
to
"Douglas G. Kilday" <fuf...@chorus.net> wrote

> In fact, let's see you compose a
> realistic archaic Latin epitaph in Saturnians, using only evidence derived
> from modern Romance vernacular texts.

Thank you for the laugh. In my reconstruction, Latin has no dative,
and the existence of a genitive is highly controversial.

Andrew Usher

unread,
Oct 8, 2004, 12:07:07 AM10/8/04
to
Miguel Carrasquer <m...@wxs.nl> wrote in message news:<qdq4m0lnjfnfkls8q...@4ax.com>...

> >In messages such as 'Re: Invention of Language', Nov 5, 1996. There
> >were many others in 1996-98, you may search likely terms in Google, if
> >you need to refresh your memory.
> >
> >Basically, the two conjectures I especially noticed were:
> >
> >1. That Etruscan is part of the IE family.
> >
> >In the past, I have always read total denials that Etruscan is related
> >to IE and even suggestions (contrary to the archaeology) that they
> >were native to Italy.
> >When I saw the comparison tables you made, I was conviced that
> >Tyrrhenian was a highly divergent branch of the IE family.
>
> I prefer to see Etruscan (and its relatives Lemnian and
> perhaps Rhaetic) as descendants of a sister language of PIE,
> the common ancestor of which I'd call "Indo-Tyrrhaenian".

OK, would you consider Anatolian to be in IE?

I would prefer to call Indo-Tyrrhenian the family; the only surviving
members of that family are IE. There are quite a few non-IE extinct
branches - we know of Anatolian and Etruscan; but there are also the
substrata of Greek, which must be Indo-Tyrrhenian.

> >2. That quite a few of the world's language families originated in the
> >Near East and spread because of the start of farming.
>
> Farming is one mechanism for language spread. The growth and
> expansion of some language families began with farming:
> Semitic, Indo-European, perhaps Dravidian. The phenomenon
> is not limited to the Near East: the spread of Chinese and
> Mayan are examples elsewhere.

Yes.

> >> I believe now as I did then that the geographical location
> >> of the Proto-Indo-European speakers was in the Early
> >> Neolithic Balkans, and that the main body of Indo-European
> >> speakers ca 5500 started to expand throughout the Northern
> >> European lowlands in what is known to archaeologists as the
> >> LBK (LinearBandKeramik) or Linear Ware culture (an earlier
> >> name is "Danubian culture"). This theory is not uniquely
> >> mine: it was advocated by several archaeologists and
> >> historians before (e.g. I basically got the idea from Colin
> >> McEvedy's Penguin Atlas of Ancient History). It got a lot
> >> of attention in Colin Renfrew's version, which was
> >> unfortunate in a way, since Renfrew managed to get almost
> >> all the linguistic arguments wrong.
> >
> >Yes. May I find online a decent summary of Renfrew's arguments? My
> >university doesn't have the book (!).
>
> You can get it from amazon used $10, new $25.

Ah. It would still be nice to read a summary rather than the entire
book.

> >> >There seems to be no source to which I could go to get his complete
> >> >argument, of which only pieces were presented here. I will thus add my
> >> >own speculation.
> >> >
> >> >It would seem to me that before the invention of agriculture (and
> >> >civilsation in general) no one would need to learn any foreign
> >> >language, and thus great linguistic diversity could arise.
> >>
> >> On the contrary: before agriculture _everybody_ had to learn
> >> one or more foreign languages, to communicate with
> >> neighbouring tribes, all of which spoke a different
> >> language. It was only with the advent of agriculture that
> >> population densities rose to such levels that most people
> >> would mostly meet other people speaking the same language in
> >> their lifetime.
> >
> >I must question your assumptions here.
> >
> >What I had in mind was fairly large areas containing tribes speaking
> >mutually intelligible dialects. They would certainly meet foreigners
> >but would not have any reason to learn their language; indeed,
> >probably the only significant interaction between them was war.
>
> Marriage, travel, exchange/trade...

If the areas were large enough these wouldn't matter. I doubt there
was any long-distance trade in the Palaeolithic, anyhow.

> >If there were not such areas, and every tribe spoke a different language,
> >how do you explain those six ancient families; as pidgins? I doubt it.
>
> I don't follow. Before the rise of farming, many languages
> must have been spoken in the area around the ancient Near
> East, just like anywhere else. When farming was "invented",
> it was taken up by some of the tribes, and their languages
> survived, at least initially (Sumerian, Hurrian, etc.), and
> perhaps not by others, who were quickly outnumbered by the
> farmers, so that we don't know what languages they spoke.

So you'd say that these language might have been very small when
agriculture arrived?

The problem with that, I think, is that agriculture would be adopted
by many tribe (since they are all facing the same pressures) rapidly
enough that one (I assume a few hundred people at most) would not have
time to expand to the size that your proto-languages must have had.
For example, we know that farming crossed into Europe by 7000 BC,
leaving only a thousand years for expansion.

> >In all of history until the 20th century most people were monolingual
> >and that is my default assumption for prehistoric times, as well.
>
> That is wrong. Most people in hunter/gatherer societies are
> multilingual.

Are? We're talking about 10,000 BC here!

> >As well, I would think that any nomadic tribe would meet several
> >hundred others in a human lifetime; do you think any speaker learned
> >the language of every one? Absurd.
>
> One learnt the language of one's own tribe, the language of
> the other moiety of one's own tribe, where applicable, and
> one learnt to make oneself understood in the languages of
> the most important nearby tribes, especially the ones that
> were a source for spouses, and/or in the local lingua
> franca, where applicable.

I don't know the word 'moiety'. What is it?

Again, this assumes the 'nearby tribes' spoke unintelligible
languages, which I assumed not to be the case above. Also, 'making
oneself understood' does not require fluency, and only one (or a few)
member of the tribe actually needs even that.

Andrew Usher

Andrew Usher

unread,
Oct 8, 2004, 12:10:54 AM10/8/04
to
"John Atkinson" <john...@bigpond.com> wrote in message news:<41q8d.14597$5O5....@news-server.bigpond.net.au>...

> > I am very suspicious that many of
> > those 'languages' are mutually intelligible dialects. Anthropologists
> > will naturally be biased toward recognising as many as possible, and
> > every tribe will insist on the uniqueness of 'their' language.
>
> Like in Australia, where there are (were, rather) around 250 "languages" by
> linguistic criteria, as opposed to seven hundred or more "political"
> languages, by which different tribes identified themselves.
>
> "Anthropologists" may, or may not, be biased like that, I wouldn't know.
> Linguists, however, don't pay too much attention to them in doing their
> counts.

What methods do they use? Surely in would be an inordinate effort to
document every tribe sufficiently to use the classic IE methods,
right?

> Foley divides the Papuan languages into "upwards of sixty distinct language
> families". That's sixty families that, like Indo-European, haven't be
> proved to be related to any other using the comparative method. In fact,
> though, many of these families are not much more divergent than the Romance
> family. Most people expect the number of "unrelated" families to end up
> considerably less than sixty once more work has been done. Various more or
> less sensible hypotheses exist, from the ridiculous "Indo-Pacific" that
> contains all of them and more, through intermediate entities like the
> Trans-New-Guinea Phylum, to groupings like Proto-Highlands, for which there
> is quite good evidence, more suggestive than probative at this stage
> however.

Well, I wouldn't know. Considering that they're all illilterate, and
considering the strange cultural dissimilation that Mr. Guy posted
about, reconstruction seems impossible.

Andrew Usher

Andrew Usher

unread,
Oct 8, 2004, 12:15:06 AM10/8/04
to
"Douglas G. Kilday" <fuf...@chorus.net> wrote in message news:<41641...@newspeer2.tds.net>...

> > One of the 'myths' is that which opposes any detailed reconstruction
> > of a proto-language. Recent works by linguists I have inspected are
> > alarmingly vague about PIE; many refuse to even say that we can
> > reconstruct the actual sounds. I do think that linguistics is a
> > science; that scientific endeavor need not be directed toward anything
> > practical or of service to other fields of study; and that detailed
> > reconstruction of proto-languages therefore may well be pursued, and
> > that a marker of sufficiently good evidence is the ability to compose
> > text in the proto-language.
>
> All right, smart guy, let's see you reconstruct archaic Latin text on the
> exclusive basis of modern Romance vernacular texts. Let's see you
> reconstruct the ablative suffix <-d>, the middle imperative suffix <-mino>,
> and subjunctives like <siem>, <edim>, <duim>. Let's see you reconstruct
> archaic Latin syntax in its entirety. Let's see you reconstruct the
> Saturnian metre. Let's see you pull Latin lexemes out of your magician's
> hat which have no reflexes in Romance. In fact, let's see you compose a
> realistic archaic Latin epitaph in Saturnians, using only evidence derived
> from modern Romance vernacular texts.

Of course, you can't. From the Romance language you can only get the
last common ancestor - 4th century colloquial Latin.

Take the noun cases, for example. In Late VL, the ablative had
disappeared and the dative and genitive were thoroughly confused. I
would therefore expect to reconstruct three cases - Nom, Acc, Dat/Gen;
this is exactly what you would get.

Further, no reconstruction will be absolutely perfect. That hardly
makes the effort worthless.

Andrew Usher

Jacques Guy

unread,
Oct 8, 2004, 5:44:57 PM10/8/04
to
Andrew Usher wrote:

> Well, I wouldn't know. Considering that they're all illilterate, and
> considering the strange cultural dissimilation that Mr. Guy posted
> about, reconstruction seems impossible.

Let me quote, again, Arthur Capell (who is he? Do a google
search for '"arthur capell" australian') at a seminar
at the Australian National University, many years ago,
even before I got my PhD. And, let's face it, that is
a bloody long time ago.

Capell was talking about the problems of comparative
Australian linguistics. "When you try to reconstruct
Proto-Australian," he said, "what do you end up with?
[a short silence here, then he raised his index finger]
_one_ word?"

That was more than thirty years ago, but things have
not changed. If we believe the archaeologists human
occupation of Australia is anything from 40,000 years
to 100,000 years old . It is hardly any surprise that
nothing should have survived of the original language.
If there was _one_ original language. Even reconstructing
a 10,000 years old language is a fool's errand.

John Atkinson

unread,
Oct 8, 2004, 6:28:54 AM10/8/04
to

"Andrew Usher" <k_over...@yahoo.com> wrote...

> "John Atkinson" <john...@bigpond.com> wrote...
>
> > AU:


> > > I am very suspicious that many of
> > > those 'languages' are mutually intelligible dialects. Anthropologists
> > > will naturally be biased toward recognising as many as possible, and
> > > every tribe will insist on the uniqueness of 'their' language.
> >
> > Like in Australia, where there are (were, rather) around 250 "languages"
by
> > linguistic criteria, as opposed to seven hundred or more "political"
> > languages, by which different tribes identified themselves.
> >
> > "Anthropologists" may, or may not, be biased like that, I wouldn't know.
> > Linguists, however, don't pay too much attention to them in doing their
> > counts.
>
> What methods do they use?

Several. None of them free of criticism. Probably the best known (and most
criticised) are the "mutual intelligibility" criterion, the "proportion of
similar sounding words" criterion, and the "the Ethnologue says" criterion.
All three have the advantage that you don't have to be fluent in the
language(s) concerned yourself to use them, and the disadvantage that the
closer you look the more contradictory conclusions you come up with.

In fact, "the number of languages spoken in X" is a fuzzy quantity, no
matter how you define it. But for the sake of handwaving discussions like
this thread, the fuzziness doesn't matter. In non-hand-waving environments,
linguists customarily use weasel words like "variety" instead.

> Surely in would be an inordinate effort to
> document every tribe sufficiently to use the classic IE methods,
> right?

I don't know what "classical IE methods" means in this context. More hot
air has been expended on defining what are "languages" in IE than any other
family. I refer you to the threads on *nglish/*cots, *anish/*orwegian (both
kinds)/*wedish, *erbian/*roatian/*osnian/*acedonian/*ulgarian, and (probably
the fuzziest of all, certainly worthy of a thread somewhere)
Hindi/Urdu/Braj/Marwari/Awadhi/Maithili/Garhwali/...

> Surely in would be an inordinate effort to
> document every tribe

Full documentation of every one of the world's 6000 languages while it's
still possible is exactly what descriptive linguists dream of having.

John.

Peter T. Daniels

unread,
Oct 8, 2004, 7:31:21 AM10/8/04
to
Andrew Usher wrote:

> > >Yes. May I find online a decent summary of Renfrew's arguments? My
> > >university doesn't have the book (!).
> >
> > You can get it from amazon used $10, new $25.
>
> Ah. It would still be nice to read a summary rather than the entire
> book.

Simple. Renfrew doesn't know anything about historical linguistics, and
repeatedly admits so, and all his suggestions about linguistic
situations (correlations with archeological evidence) in the distant
past are simply not viable.

Mallory's book came out shortly thereafter, just enough later that he
was able to discuss Renfrew in a few lengthy footnotes. That's *In
Search of the Indo-Europeans* (Thames & Hudson, 1989), and though
somewhat dated it remains the standard treatment.

Peter T. Daniels

unread,
Oct 8, 2004, 7:39:05 AM10/8/04
to
Andrew Usher wrote:
>
> Miguel Carrasquer <m...@wxs.nl> wrote in message news:<qdq4m0lnjfnfkls8q...@4ax.com>...

> > I prefer to see Etruscan (and its relatives Lemnian and


> > perhaps Rhaetic) as descendants of a sister language of PIE,
> > the common ancestor of which I'd call "Indo-Tyrrhaenian".
>
> OK, would you consider Anatolian to be in IE?

What's the basis of your knowledge of Anatolian -- are you qualified to
formulate an opinion of your own?

> I would prefer to call Indo-Tyrrhenian the family; the only surviving
> members of that family are IE. There are quite a few non-IE extinct
> branches - we know of Anatolian and Etruscan; but there are also the
> substrata of Greek, which must be Indo-Tyrrhenian.

What's the knowledge-base for your assertions about IE and its closest
relatives?

> > >What I had in mind was fairly large areas containing tribes speaking
> > >mutually intelligible dialects. They would certainly meet foreigners
> > >but would not have any reason to learn their language; indeed,
> > >probably the only significant interaction between them was war.
> >
> > Marriage, travel, exchange/trade...
>
> If the areas were large enough these wouldn't matter. I doubt there
> was any long-distance trade in the Palaeolithic, anyhow.

Individual persons may not have crossed all of Eurasia, but materials
are found in one site that originated thousands of miles away (obsidian,
I think, is one such). Either it passed from hand to hand across the
continent, or there really were traveling salesmen back then.

> > >In all of history until the 20th century most people were monolingual
> > >and that is my default assumption for prehistoric times, as well.
> >
> > That is wrong. Most people in hunter/gatherer societies are
> > multilingual.
>
> Are? We're talking about 10,000 BC here!

Whatever is known about hunter-gatherer lifeways comes from observation
of 20th-century hunter-gatherers. And most people outside the Anglophone
world are multilingual.

> > >As well, I would think that any nomadic tribe would meet several
> > >hundred others in a human lifetime; do you think any speaker learned
> > >the language of every one? Absurd.
> >
> > One learnt the language of one's own tribe, the language of
> > the other moiety of one's own tribe, where applicable, and
> > one learnt to make oneself understood in the languages of
> > the most important nearby tribes, especially the ones that
> > were a source for spouses, and/or in the local lingua
> > franca, where applicable.
>
> I don't know the word 'moiety'. What is it?

The other half of one's tribe, from which one may take a spouse.

> Again, this assumes the 'nearby tribes' spoke unintelligible
> languages, which I assumed not to be the case above. Also, 'making
> oneself understood' does not require fluency, and only one (or a few)
> member of the tribe actually needs even that.

Why would you assume "nearby tribes"' languages would be intelligible?

Miguel Carrasquer

unread,
Oct 9, 2004, 3:46:53 PM10/9/04
to
On 7 Oct 2004 21:07:07 -0700, k_over...@yahoo.com (Andrew
Usher) wrote:

>Miguel Carrasquer <m...@wxs.nl> wrote in message news:<qdq4m0lnjfnfkls8q...@4ax.com>...
>
>> >In messages such as 'Re: Invention of Language', Nov 5, 1996. There
>> >were many others in 1996-98, you may search likely terms in Google, if
>> >you need to refresh your memory.
>> >
>> >Basically, the two conjectures I especially noticed were:
>> >
>> >1. That Etruscan is part of the IE family.
>> >
>> >In the past, I have always read total denials that Etruscan is related
>> >to IE and even suggestions (contrary to the archaeology) that they
>> >were native to Italy.
>> >When I saw the comparison tables you made, I was conviced that
>> >Tyrrhenian was a highly divergent branch of the IE family.
>>
>> I prefer to see Etruscan (and its relatives Lemnian and
>> perhaps Rhaetic) as descendants of a sister language of PIE,
>> the common ancestor of which I'd call "Indo-Tyrrhaenian".
>
>OK, would you consider Anatolian to be in IE?

Yes.

>> >What I had in mind was fairly large areas containing tribes speaking
>> >mutually intelligible dialects. They would certainly meet foreigners
>> >but would not have any reason to learn their language; indeed,
>> >probably the only significant interaction between them was war.
>>
>> Marriage, travel, exchange/trade...
>
>If the areas were large enough these wouldn't matter. I doubt there
>was any long-distance trade in the Palaeolithic, anyhow.

It doesn't have to be long-distance.

>> >If there were not such areas, and every tribe spoke a different language,
>> >how do you explain those six ancient families; as pidgins? I doubt it.
>>
>> I don't follow. Before the rise of farming, many languages
>> must have been spoken in the area around the ancient Near
>> East, just like anywhere else. When farming was "invented",
>> it was taken up by some of the tribes, and their languages
>> survived, at least initially (Sumerian, Hurrian, etc.), and
>> perhaps not by others, who were quickly outnumbered by the
>> farmers, so that we don't know what languages they spoke.
>
>So you'd say that these language might have been very small when
>agriculture arrived?
>
>The problem with that, I think, is that agriculture would be adopted
>by many tribe (since they are all facing the same pressures) rapidly
>enough that one (I assume a few hundred people at most) would not have
>time to expand to the size that your proto-languages must have had.

In the area where farming was "invented" (in the Near-East:
the areas where wheat and barley were indigenous), one would
expect to see, at least initially, many language families
(Sumerian, Hurrian, Elamite, Semitic, Indo-European,
Kartvelian). Groups at the periphery (IE, Semitic) can then
spread out into areas where the technology/crops are not
indigenous and/or the climate is different. The locals have
no time to adopt the technology and are numerically
overwhelmed, until for some reason (geography, climate), the
advance stops.

Andrew Usher

unread,
Oct 10, 2004, 2:43:50 AM10/10/04
to
Miguel Carrasquer <m...@wxs.nl> wrote in message news:<6hfgm0lmrbn822gu8...@4ax.com>...

> >> >1. That Etruscan is part of the IE family.
> >> >
> >> >In the past, I have always read total denials that Etruscan is related
> >> >to IE and even suggestions (contrary to the archaeology) that they
> >> >were native to Italy.
> >> >When I saw the comparison tables you made, I was conviced that
> >> >Tyrrhenian was a highly divergent branch of the IE family.
> >>
> >> I prefer to see Etruscan (and its relatives Lemnian and
> >> perhaps Rhaetic) as descendants of a sister language of PIE,
> >> the common ancestor of which I'd call "Indo-Tyrrhaenian".
> >
> >OK, would you consider Anatolian to be in IE?
>
> Yes.

There doesn't seem to be good reason for treating them differently, as
both split off before what we call PIE.

Let's see, I will give my brief organisation (sorry for the
formatting, but I can't do better in pure ASCII):

- Indo-Tyrrhenian
> - Tyrrhenian (all extinct)
> > - Etruscan
> > - and others
> - Anatolian (all extinct)
> > - Hittite
> > - and others
> - Indo-European
> > - Tocharian (all extinct)
> > - Germanic
> > - Albanian
> > - Italo-Celtic
> > > - Celtic
> > > - Italic (all extinct)
> > > - Latin and Romance
> > - Balto-Slavic
> > > - Baltic
> > > - Slavic
> > - Indo-Greek
> > > - Indo-Iranian
> > > > - Indic
> > > > - Iranian
> > > - Greek

Above, I have attempted to order the families by when they diverged
from their parent. I haven't given any exact dates, but you could
insert them, based on the chronology you have given already.

> >So you'd say that these language might have been very small when
> >agriculture arrived?
> >
> >The problem with that, I think, is that agriculture would be adopted
> >by many tribe (since they are all facing the same pressures) rapidly
> >enough that one (I assume a few hundred people at most) would not have
> >time to expand to the size that your proto-languages must have had.
>
> In the area where farming was "invented" (in the Near-East:
> the areas where wheat and barley were indigenous), one would
> expect to see, at least initially, many language families
> (Sumerian, Hurrian, Elamite, Semitic, Indo-European,
> Kartvelian). Groups at the periphery (IE, Semitic) can then
> spread out into areas where the technology/crops are not
> indigenous and/or the climate is different. The locals have
> no time to adopt the technology and are numerically
> overwhelmed, until for some reason (geography, climate), the
> advance stops.

You did not really respond to my question above: do you hold that the
languages were spoken only by a single tribe when agriculture came
along, or over a larger area?

You are certainly right that farming started there only because wheat
and other grains were found natively; certainly not because the
climate is optimal (indeed, irrigation is almost required over much of
the area).

Andrew Usher

grapheus

unread,
Oct 10, 2004, 7:41:45 AM10/10/04
to
Miguel Carrasquer <m...@wxs.nl> wrote in message news:<qdq4m0lnjfnfkls8q...@4ax.com>...

> On 4 Oct 2004 20:58:35 -0700, k_over...@yahoo.com (Andrew
> Usher) wrote:
>
> >Miguel Carrasquer <m...@wxs.nl> wrote in message news:<89b3m097ccems59m5...@4ax.com>...
> >
> >> >I have become interested in the origins of Indoeuropean language, and
> >> >as I looked at it I thought that linguistic dogmas were preventing the
> >> >truth from being researched properly. I began to create my own theory
> >> >(without having studied enough history, I don't think) but abandoned
> >> >it some time ago when I discoved the remarkable theory expounded by
> >> >Miguel Carrasquer in this group from 1996-98.
> >>
> >> Refresh my mind, what remarkable theory was that?
> >
> >In messages such as 'Re: Invention of Language', Nov 5, 1996. There
> >were many others in 1996-98, you may search likely terms in Google, if
> >you need to refresh your memory.
> >
> >Basically, the two conjectures I especially noticed were:
> >
> >1. That Etruscan is part of the IE family.
> >
> >In the past, I have always read total denials that Etruscan is related
> >to IE and even suggestions (contrary to the archaeology) that they
> >were native to Italy.
> >When I saw the comparison tables you made, I was conviced that
> >Tyrrhenian was a highly divergent branch of the IE family.

Hey, guys !.. You are rejoining the "Kretschmerian Theory" concerning
the "Proto-IE" !.. A VERY GOOD MOVE...


>
> I prefer to see Etruscan (and its relatives Lemnian and
> perhaps Rhaetic) as descendants of a sister language of PIE,
> the common ancestor of which I'd call "Indo-Tyrrhaenian".

1)- It's a question of DEFINITION concerning what is the "proto-IE".
Why not accepting the "Kretschmerian Definition" about this word ?
2)- Why "Indo-" ???? As far as I know, there is no obvious links
between Tyrrhaenian" and Sanskrit. But there are some between
Tyrrhenian and Greek...

grapheus

Peter T. Daniels

unread,
Oct 10, 2004, 8:19:14 AM10/10/04
to
Andrew Usher wrote:
>
> Miguel Carrasquer <m...@wxs.nl> wrote in message news:<6hfgm0lmrbn822gu8...@4ax.com>...
>
> > >> >1. That Etruscan is part of the IE family.
> > >> >
> > >> >In the past, I have always read total denials that Etruscan is related
> > >> >to IE and even suggestions (contrary to the archaeology) that they
> > >> >were native to Italy.
> > >> >When I saw the comparison tables you made, I was conviced that
> > >> >Tyrrhenian was a highly divergent branch of the IE family.
> > >>
> > >> I prefer to see Etruscan (and its relatives Lemnian and
> > >> perhaps Rhaetic) as descendants of a sister language of PIE,
> > >> the common ancestor of which I'd call "Indo-Tyrrhaenian".
> > >
> > >OK, would you consider Anatolian to be in IE?
> >
> > Yes.
>
> There doesn't seem to be good reason for treating them differently, as
> both split off before what we call PIE.

The only reason for treating them as the same was if they exhibited
shared morphological innovations, i.e. were subgroupable, but since
"Indo-Tyrrhenian" is pure fantasy, there isn't the slightest way of
discovering whether there are any. All that can be said is that the
Anatolian family fits nicely into IE as it had been known before (just
as Potawotami fitted nicely into Proto-Central Algonquian); and that
there are reasons, fully comprehesible only to Anatolianists, for
denying that it's a separate family justifying a label "Indo-Hittite."

Nor, BTW, is it legitimate to suggest that families "diverged from their
parent" one by one. The subgrouping within IE is determinable -- and it
doesn't begin with centum/satem, either.

Why do you imagine Latin isn't part of Italic?

Brian M. Scott

unread,
Oct 10, 2004, 12:08:52 PM10/10/04
to
On 9 Oct 2004 23:43:50 -0700, Andrew Usher
<k_over...@yahoo.com> wrote in
<news:6e197594.04100...@posting.google.com> in
sci.lang:

> Miguel Carrasquer <m...@wxs.nl> wrote in message news:<6hfgm0lmrbn822gu8...@4ax.com>...

>>>>>1. That Etruscan is part of the IE family.

>>>>>In the past, I have always read total denials that Etruscan is related
>>>>>to IE and even suggestions (contrary to the archaeology) that they
>>>>>were native to Italy.
>>>>>When I saw the comparison tables you made, I was conviced that
>>>>>Tyrrhenian was a highly divergent branch of the IE family.

>>>> I prefer to see Etruscan (and its relatives Lemnian and
>>>> perhaps Rhaetic) as descendants of a sister language of PIE,
>>>> the common ancestor of which I'd call "Indo-Tyrrhaenian".

>>>OK, would you consider Anatolian to be in IE?

>> Yes.

> There doesn't seem to be good reason for treating them differently, as
> both split off before what we call PIE.

There is a very obvious reason for treating them
differently: the Anatolian languages fit readily into IE as
evidenced by the non-Anatolian branches, and Etruscan
doesn't.

> Let's see, I will give my brief organisation (sorry for the
> formatting, but I can't do better in pure ASCII):

[...]

>>> - Italo-Celtic
>>> > - Celtic
>>> > - Italic (all extinct)
>>> > - Latin and Romance

Latin is Italic. And last I knew, the existence of an
Italo-Celtic node was still up in the air.

[...]

> Above, I have attempted to order the families by when they
> diverged from their parent.

That's not a good way of thinking about it; after a split
there are only daughters, not daughter and parent.

[...]

>>>So you'd say that these language might have been very small when
>>>agriculture arrived?

>>>The problem with that, I think, is that agriculture would be adopted
>>>by many tribe (since they are all facing the same pressures) rapidly
>>>enough that one (I assume a few hundred people at most) would not have
>>>time to expand to the size that your proto-languages must have had.

>> In the area where farming was "invented" (in the Near-East:
>> the areas where wheat and barley were indigenous), one would
>> expect to see, at least initially, many language families
>> (Sumerian, Hurrian, Elamite, Semitic, Indo-European,
>> Kartvelian). Groups at the periphery (IE, Semitic) can then
>> spread out into areas where the technology/crops are not
>> indigenous and/or the climate is different. The locals have
>> no time to adopt the technology and are numerically
>> overwhelmed, until for some reason (geography, climate), the
>> advance stops.

> You did not really respond to my question above:

Of course he did: he offered an explanation of why IE and
Semitic could in fact have expanded to a size that you said
was impossible.

[...]

Brian

Peter T. Daniels

unread,
Oct 10, 2004, 12:15:03 PM10/10/04
to
Brian M. Scott wrote:

> Latin is Italic. And last I knew, the existence of an
> Italo-Celtic node was still up in the air.

Eric Hamp says epigraphic data from the past forty years make it highly
likely -- a change from the last careful investigation of the question
ca. 1960.

Brian M. Scott

unread,
Oct 10, 2004, 1:23:30 PM10/10/04
to
On Sun, 10 Oct 2004 16:15:03 GMT, "Peter T. Daniels"
<gram...@worldnet.att.net> wrote in
<news:416960...@worldnet.att.net> in sci.lang:

> Brian M. Scott wrote:

>> Latin is Italic. And last I knew, the existence of an
>> Italo-Celtic node was still up in the air.

> Eric Hamp says epigraphic data from the past forty years make it highly
> likely -- a change from the last careful investigation of the question
> ca. 1960.

Thanks.

Brian

Neeraj Mathur

unread,
Oct 10, 2004, 7:53:03 PM10/10/04
to

"Peter T. Daniels" <gram...@worldnet.att.net> wrote in message
news:416928...@worldnet.att.net...

> Why do you imagine Latin isn't part of Italic?

Does the latest consensus opinion treat Latin-Faliscan and Oscan-Umbrian as
separate, independent branches of IE that show areal contact? Or are we
still saying that these are two subgroups of a common Italic? Or does
controversy still hold?

Neeraj Mathur


Peter T. Daniels

unread,
Oct 11, 2004, 7:06:26 AM10/11/04
to

Is there a "consensus"?; There are more than four languages involved;
and see Baldi, Foundations of Latin, for an overview of many proposals.

Miguel Carrasquer

unread,
Oct 11, 2004, 3:46:13 PM10/11/04
to
On 9 Oct 2004 23:43:50 -0700, k_over...@yahoo.com (Andrew
Usher) wrote:

>Miguel Carrasquer <m...@wxs.nl> wrote in message news:<6hfgm0lmrbn822gu8...@4ax.com>...
>
>> >> >1. That Etruscan is part of the IE family.
>> >> >
>> >> >In the past, I have always read total denials that Etruscan is related
>> >> >to IE and even suggestions (contrary to the archaeology) that they
>> >> >were native to Italy.
>> >> >When I saw the comparison tables you made, I was conviced that
>> >> >Tyrrhenian was a highly divergent branch of the IE family.
>> >>
>> >> I prefer to see Etruscan (and its relatives Lemnian and
>> >> perhaps Rhaetic) as descendants of a sister language of PIE,
>> >> the common ancestor of which I'd call "Indo-Tyrrhaenian".
>> >
>> >OK, would you consider Anatolian to be in IE?
>>
>> Yes.
>
>There doesn't seem to be good reason for treating them differently, as
>both split off before what we call PIE.

No, Anatolian split off _after_ what I call PIE.

What we know about Etruscan is pitifully little. Etruscan
can therefore play no significant part in the reconstruction
of PIE.

Andrew Usher

unread,
Oct 12, 2004, 1:06:50 AM10/12/04
to
Miguel Carrasquer <m...@wxs.nl> wrote in message news:<ejolm09rvrtmicu59...@4ax.com>...

> >There doesn't seem to be good reason for treating them differently, as
> >both split off before what we call PIE.
>
> No, Anatolian split off _after_ what I call PIE.

OK, after your PIE. But the normal reconstructions treat it as if all
languages diverged at the same time, so that PIE will be after the
Hittite split. In fact, it will have some purely Indo-Greek features,
especially if you assume (as was the older practice) that
morphological complexity is only lost.

> What we know about Etruscan is pitifully little. Etruscan
> can therefore play no significant part in the reconstruction
> of PIE.

Indeed this is unfortunate; one would really hope that there would be
more, thanks to them having lived with the Romans ffor centuries; but
the Romans don't seem to have been interested in documenting other
people's speech - I don't believe we have anything by them on the
Celtic or Germanic languages.

Unless this changes, I do not think we will be able to go back to
Proto-Indo-Tyrrhenian (comprehensively, anyway).

Andrew Usher

Andrew Usher

unread,
Oct 12, 2004, 1:18:20 AM10/12/04
to
"Peter T. Daniels" <gram...@worldnet.att.net> wrote in message news:<416928...@worldnet.att.net>...

(Anatolian and Tyrrhenian)

> > There doesn't seem to be good reason for treating them differently, as
> > both split off before what we call PIE.
>
> The only reason for treating them as the same was if they exhibited
> shared morphological innovations, i.e. were subgroupable, but since
> "Indo-Tyrrhenian" is pure fantasy, there isn't the slightest way of
> discovering whether there are any. All that can be said is that the
> Anatolian family fits nicely into IE as it had been known before (just
> as Potawotami fitted nicely into Proto-Central Algonquian); and that
> there are reasons, fully comprehesible only to Anatolianists, for
> denying that it's a separate family justifying a label "Indo-Hittite."

I did not say that Anatolian and Tyrrhenian were a subgroup, but they
they were both related to IE. As far as the Indo-Tyrrhenian
hypothesis, I reproduce here Mr. Carrasquer's table of
correspondences:

PIE Etruscan

* *-m -n (pron.) 'accusative'
* *-s -s(i) 'genitive'
*(Hitt.) *-l (pron.) -l(a) 'genitive'
* *-i -i 'locative/dative'
* *-nt- -nth 'pres. ptc.'
* *-to- -tha- 'pf. ptc.'
* *-kwe -c 'and'
*(Hitt.) *-ma -m 'and, but'
* *-bhi -pi 'for, at, ...'
* *me- (nom.) mi 'I, me'
* (acc.) mini
* *ko- i-ca 'this, that'
* *to- i-ta 'this, that'
* *'ag- ac- 'to drive, to do'
* *hanti- hanthe 'before'
* *leudh- laut-n 'free(dman), people, family'
* *nepot- nefts 'nephew, grandson'
* *sed- sut(h)-,sat(h)-'to sit, to set'
* *dei-n- tin 'day'
* *dei-u- tiu(r) 'light > moon, month'
* *doh- tur 'to give' (cf. Grk. do:-r-on)
* *dem-/dom- tmia 'house, temple'
* tham- 'to settle, to establish'
* *sahu-l- usil 'sun'

The basic noun cases are particularly interesting, as they should
never be borrowed. Even this table would not be convincing, were there
not geographical reasons for the connection. Of course, we really need
to decipher more Etruscan to be quite sure.

Obviously there is no fixed 'parent language' to diverge from, but as
a manner of speaking, it is certainly legitimate to talk about the
date at which one family separated from all its relatives.

(And I did not use centum/satem either. Palatalisation before front
vowels is rather common; there is no reason it could not have happened
independently more than once.)

> Why do you imagine Latin isn't part of Italic?

Again, I am not sure. But it is known that there were more than one
'invasion' of Italy by IE speakers, and I don't think there ever was a
reason other than geography to group them together. (True, both tend
to be phonologically conservative, but so does Greek, and it isn't
Italic.)

Andrew Usher

Andrew Usher

unread,
Oct 12, 2004, 1:25:40 AM10/12/04
to
"John Atkinson" <john...@bigpond.com> wrote in message news:<GZt9d.18246$5O5....@news-server.bigpond.net.au>...

> > What methods do they use?
>
> Several. None of them free of criticism. Probably the best known (and most
> criticised) are the "mutual intelligibility" criterion, the "proportion of
> similar sounding words" criterion, and the "the Ethnologue says" criterion.
> All three have the advantage that you don't have to be fluent in the
> language(s) concerned yourself to use them, and the disadvantage that the
> closer you look the more contradictory conclusions you come up with.
>
> In fact, "the number of languages spoken in X" is a fuzzy quantity, no
> matter how you define it. But for the sake of handwaving discussions like
> this thread, the fuzziness doesn't matter. In non-hand-waving environments,
> linguists customarily use weasel words like "variety" instead.

Yes, it is. The point being made here was that these islands have a
far greater degree of linguistic diversity than Europe and Asia do.

> > Surely in would be an inordinate effort to
> > document every tribe sufficiently to use the classic IE methods,
> > right?
>
> I don't know what "classical IE methods" means in this context. More hot
> air has been expended on defining what are "languages" in IE than any other
> family. I refer you to the threads on *nglish/*cots, *anish/*orwegian (both
> kinds)/*wedish, *erbian/*roatian/*osnian/*acedonian/*ulgarian, and (probably
> the fuzziest of all, certainly worthy of a thread somewhere)
> Hindi/Urdu/Braj/Marwari/Awadhi/Maithili/Garhwali/...

The methods of comparative linguistic were first perfected with IE
languages, that is all.

That kind of dispute very commonly arises when politics and
nationalism get involved. The truth is that many language families are
merely dialect continua (or would be if governments did not so
aggresively promote national languages).

> > Surely in would be an inordinate effort to
> > document every tribe
>
> Full documentation of every one of the world's 6000 languages while it's
> still possible is exactly what descriptive linguists dream of having.

It was Mr. Guy's assertion that this had been done for New-Guinea and
Vanuatu, was it not?

Andrew Usher

Jacques Guy

unread,
Oct 12, 2004, 7:42:25 PM10/12/04
to
Andrew Usher wrote:
>
> "John Atkinson" <john...@bigpond.com> wrote in message news:<GZt9d.18246$5O5....@news-server.bigpond.net.au>...

> > Full documentation of every one of the world's 6000 languages while it's


> > still possible is exactly what descriptive linguists dream of having.

> It was Mr. Guy's assertion that this had been done for New-Guinea and
> Vanuatu, was it not?

No. Most of those languages have been _sampled_, using
extended Swadesh word lists. Full documentation is
something else. Very few have been fully documented.

Peter T. Daniels

unread,
Oct 12, 2004, 9:03:24 AM10/12/04
to
Andrew Usher wrote:
>
> Miguel Carrasquer <m...@wxs.nl> wrote in message news:<ejolm09rvrtmicu59...@4ax.com>...
>
> > >There doesn't seem to be good reason for treating them differently, as
> > >both split off before what we call PIE.
> >
> > No, Anatolian split off _after_ what I call PIE.
>
> OK, after your PIE. But the normal reconstructions treat it as if all
> languages diverged at the same time,

That's the second time you've said that, and it still isn't true. In
fact, since you've already been apprised otherwise, it's a lie.

> so that PIE will be after the
> Hittite split. In fact, it will have some purely Indo-Greek features,
> especially if you assume (as was the older practice) that
> morphological complexity is only lost.
>
> > What we know about Etruscan is pitifully little. Etruscan
> > can therefore play no significant part in the reconstruction
> > of PIE.
>
> Indeed this is unfortunate; one would really hope that there would be
> more, thanks to them having lived with the Romans ffor centuries; but
> the Romans don't seem to have been interested in documenting other
> people's speech - I don't believe we have anything by them on the
> Celtic or Germanic languages.
>
> Unless this changes, I do not think we will be able to go back to
> Proto-Indo-Tyrrhenian (comprehensively, anyway).

Indeed, to use such a name is self-delusion.

Peter T. Daniels

unread,
Oct 12, 2004, 9:08:44 AM10/12/04
to
Andrew Usher wrote:
>
> "Peter T. Daniels" <gram...@worldnet.att.net> wrote in message news:<416928...@worldnet.att.net>...
>
> (Anatolian and Tyrrhenian)

No, it was clear from the context that you were referring to Anatolian
and IE.

> > > There doesn't seem to be good reason for treating them differently, as
> > > both split off before what we call PIE.
> >
> > The only reason for treating them as the same was if they exhibited
> > shared morphological innovations, i.e. were subgroupable, but since
> > "Indo-Tyrrhenian" is pure fantasy, there isn't the slightest way of
> > discovering whether there are any. All that can be said is that the
> > Anatolian family fits nicely into IE as it had been known before (just
> > as Potawotami fitted nicely into Proto-Central Algonquian); and that
> > there are reasons, fully comprehesible only to Anatolianists, for
> > denying that it's a separate family justifying a label "Indo-Hittite."
>
> I did not say that Anatolian and Tyrrhenian were a subgroup, but they

I did not suggest that you did.

> they were both related to IE. As far as the Indo-Tyrrhenian
> hypothesis, I reproduce here Mr. Carrasquer's table of
> correspondences:

Why is Miguel's interpretation of Etruscan any more valid than anyone
else's? How does the scheme you presented (which I snipped) compare with
Georgiev's, for instance?

This does not happen. Populations split up, and the resulting
populations split up, and so on. You don't get one family vs. all the
others.

> (And I did not use centum/satem either. Palatalisation before front

That's the problem. You didn't use _any_ subgrouping, so your model is
absurd.

> vowels is rather common; there is no reason it could not have happened
> independently more than once.)
>
> > Why do you imagine Latin isn't part of Italic?
>
> Again, I am not sure. But it is known that there were more than one
> 'invasion' of Italy by IE speakers, and I don't think there ever was a
> reason other than geography to group them together. (True, both tend
> to be phonologically conservative, but so does Greek, and it isn't
> Italic.)

Invasion shminvasion, what are the linguistic features of Latin that
lead you to suppose it's not an Italic language?

Andrew Usher

unread,
Oct 13, 2004, 1:41:01 AM10/13/04
to
Jacques Guy <jg...@alphalink.com.au> wrote in message

> > > Full documentation of every one of the world's 6000 languages while it's
> > > still possible is exactly what descriptive linguists dream of having.
>
> > It was Mr. Guy's assertion that this had been done for New-Guinea and
> > Vanuatu, was it not?
>
> No. Most of those languages have been _sampled_, using
> extended Swadesh word lists. Full documentation is
> something else. Very few have been fully documented.

Ah, pardon me. Would 'full documentation' mean a grammar and a dictionary?

Andrew Usher

benlizross

unread,
Oct 13, 2004, 2:58:08 AM10/13/04
to

The "comprehensive survey" I mentioned earlier as having been done in
the 1970s consisted of just such word lists as Jacques refers to. It was
comprehensive in that no large areas remained terra incognita, though it
has become clear that the odd language was missed here and there. Even
using generous definitions, the number of languages for which we have a
published grammar _or_ dictionary would be less than 1/3 of the total.

Ross Clark

Jacques Guy

unread,
Oct 13, 2004, 7:17:39 PM10/13/04
to
Andrew Usher wrote:

> Ah, pardon me. Would 'full documentation' mean a grammar and a dictionary?

Once upon a time I would have answered "yes". Nowadays I am more
inclined
to "a bilingual corpus, from which a grammar and a dictionary can be
derived". For let's face it: how do you write a description of a
language
and a dictionary, if not from a corpus of texts ("text" as in "spoken
text"
in the case of these languages)? The grammar and the dictionary are only
an individual linguist's interpretation of the language, extracted from
a corpus. Without the evidence--the corpus--I do not think they are
worth
much. Better than nothing of course, but not awfully much.

Andrew Usher

unread,
Oct 13, 2004, 1:48:14 AM10/13/04
to
"Peter T. Daniels" <gram...@worldnet.att.net> wrote in message news:<416BD7...@worldnet.att.net>...

> > (Anatolian and Tyrrhenian)
>
> No, it was clear from the context that you were referring to Anatolian
> and IE.

The context was right below: I said '... both split off before what we
call PIE.' That makes it clear that I did not mean IE.

> > I did not say that Anatolian and Tyrrhenian were a subgroup, but they
>
> I did not suggest that you did.
>
> > they were both related to IE. As far as the Indo-Tyrrhenian
> > hypothesis, I reproduce here Mr. Carrasquer's table of
> > correspondences:
>
> Why is Miguel's interpretation of Etruscan any more valid than anyone
> else's? How does the scheme you presented (which I snipped) compare with
> Georgiev's, for instance?

Why not?

What is Georgiev's work? What did he conclude Etruscan was related to
- Semitic? Basque?

> > > > Above, I have attempted to order the families by when they diverged
> > > > from their parent. I haven't given any exact dates, but you could
> > > > insert them, based on the chronology you have given already.
> > >
> > > Nor, BTW, is it legitimate to suggest that families "diverged from their
> > > parent" one by one. The subgrouping within IE is determinable -- and it
> > > doesn't begin with centum/satem, either.
> >
> > Obviously there is no fixed 'parent language' to diverge from, but as
> > a manner of speaking, it is certainly legitimate to talk about the
> > date at which one family separated from all its relatives.
>
> This does not happen. Populations split up, and the resulting
> populations split up, and so on. You don't get one family vs. all the
> others.

Are you really saying that one can't make a timeline of language
developement? That one could not say, for example, that English split
from Germanic about 2,000 years ago?

> > (And I did not use centum/satem either. Palatalisation before front
>
> That's the problem. You didn't use _any_ subgrouping, so your model is
> absurd.

I did use a subgrouping, and ordered them according to the rule I
already gave.

> > vowels is rather common; there is no reason it could not have happened
> > independently more than once.)
> >
> > > Why do you imagine Latin isn't part of Italic?
> >
> > Again, I am not sure. But it is known that there were more than one
> > 'invasion' of Italy by IE speakers, and I don't think there ever was a
> > reason other than geography to group them together. (True, both tend
> > to be phonologically conservative, but so does Greek, and it isn't
> > Italic.)
>
> Invasion shminvasion, what are the linguistic features of Latin that
> lead you to suppose it's not an Italic language?

What are the features that lead you to suppose it is?

Now I am not an authority on IE linguistics, obviously, but it appears
that you are claiming to be; in that case, I trust you could explain
some of your pronouncements.

Andrew Usher

Andrew Usher

unread,
Oct 13, 2004, 1:53:32 AM10/13/04
to
"Peter T. Daniels" <gram...@worldnet.att.net> wrote in message news:<416BD6...@worldnet.att.net>...

> > > >There doesn't seem to be good reason for treating them differently, as
> > > >both split off before what we call PIE.
> > >
> > > No, Anatolian split off _after_ what I call PIE.
> >
> > OK, after your PIE. But the normal reconstructions treat it as if all
> > languages diverged at the same time,
>
> That's the second time you've said that, and it still isn't true. In
> fact, since you've already been apprised otherwise, it's a lie.

In the other reply you stated that we can't even talk about when
languages diverged from each other. The above statement would
contradict that, would it not?

I said, to give a fuller explanation, that the standard reconstruction
treats all of the traditional subfamilies as significant to the
structure of PIE.

> > so that PIE will be after the
> > Hittite split. In fact, it will have some purely Indo-Greek features,
> > especially if you assume (as was the older practice) that
> > morphological complexity is only lost.
> >
> > > What we know about Etruscan is pitifully little. Etruscan
> > > can therefore play no significant part in the reconstruction
> > > of PIE.
> >
> > Indeed this is unfortunate; one would really hope that there would be
> > more, thanks to them having lived with the Romans ffor centuries; but
> > the Romans don't seem to have been interested in documenting other
> > people's speech - I don't believe we have anything by them on the
> > Celtic or Germanic languages.
> >
> > Unless this changes, I do not think we will be able to go back to
> > Proto-Indo-Tyrrhenian (comprehensively, anyway).
>
> Indeed, to use such a name is self-delusion.

Where do you think, then, Etruscan came from?

Andrew Usher

Peter T. Daniels

unread,
Oct 13, 2004, 8:42:58 AM10/13/04
to
Andrew Usher wrote:
>
> "Peter T. Daniels" <gram...@worldnet.att.net> wrote in message news:<416BD7...@worldnet.att.net>...
>
> > > (Anatolian and Tyrrhenian)
> >
> > No, it was clear from the context that you were referring to Anatolian
> > and IE.
>
> The context was right below: I said '... both split off before what we
> call PIE.' That makes it clear that I did not mean IE.

That may be what you meant, but it isn't what you said.

> > > I did not say that Anatolian and Tyrrhenian were a subgroup, but they
> >
> > I did not suggest that you did.
> >
> > > they were both related to IE. As far as the Indo-Tyrrhenian
> > > hypothesis, I reproduce here Mr. Carrasquer's table of
> > > correspondences:
> >
> > Why is Miguel's interpretation of Etruscan any more valid than anyone
> > else's? How does the scheme you presented (which I snipped) compare with
> > Georgiev's, for instance?
>
> Why not?

Because the data are inadequate and far from securely interpreted.

> What is Georgiev's work? What did he conclude Etruscan was related to
> - Semitic? Basque?

Indo-European.

The third edition of his book is in English; the first two were in
Bulgarian and IIRC Russian -- all were published in Sofia.

> > > > > Above, I have attempted to order the families by when they diverged
> > > > > from their parent. I haven't given any exact dates, but you could
> > > > > insert them, based on the chronology you have given already.
> > > >
> > > > Nor, BTW, is it legitimate to suggest that families "diverged from their
> > > > parent" one by one. The subgrouping within IE is determinable -- and it
> > > > doesn't begin with centum/satem, either.
> > >
> > > Obviously there is no fixed 'parent language' to diverge from, but as
> > > a manner of speaking, it is certainly legitimate to talk about the
> > > date at which one family separated from all its relatives.
> >
> > This does not happen. Populations split up, and the resulting
> > populations split up, and so on. You don't get one family vs. all the
> > others.
>
> Are you really saying that one can't make a timeline of language
> developement? That one could not say, for example, that English split
> from Germanic about 2,000 years ago?

Of course not! You don't know much about the history of England, do
you? (But even if you'd gotten the date right, "English" didn't "split
from" the rest of an undifferentiated "Germanic."

> > > (And I did not use centum/satem either. Palatalisation before front
> >
> > That's the problem. You didn't use _any_ subgrouping, so your model is
> > absurd.
>
> I did use a subgrouping, and ordered them according to the rule I
> already gave.

Rule?

There are no subgroupings in the list you provided; all the family-level
groups are at the same indentation with no groupings indicated.

> > > vowels is rather common; there is no reason it could not have happened
> > > independently more than once.)
> > >
> > > > Why do you imagine Latin isn't part of Italic?
> > >
> > > Again, I am not sure. But it is known that there were more than one
> > > 'invasion' of Italy by IE speakers, and I don't think there ever was a
> > > reason other than geography to group them together. (True, both tend
> > > to be phonologically conservative, but so does Greek, and it isn't
> > > Italic.)
> >
> > Invasion shminvasion, what are the linguistic features of Latin that
> > lead you to suppose it's not an Italic language?
>
> What are the features that lead you to suppose it is?

I'm not a Latinist. I can't list the features. But I have never seen a
work on the Italic languages that suggested Latin isn't one of them.
Latinists argue about the subdivision within Italic, but they don't
exclude it from it.

> Now I am not an authority on IE linguistics, obviously, but it appears
> that you are claiming to be; in that case, I trust you could explain
> some of your pronouncements.

I can only refer you to the standard works.

Jacques Guy

unread,
Oct 14, 2004, 1:56:26 AM10/14/04
to
Andrew Usher wrote:

[snip: too confusing]

> In the other reply you [PT Daniels] stated that we can't even talk about when


> languages diverged from each other. The above statement would
> contradict that, would it not?

The keyword is "when". If "when" is meant in terms of calendar time,
no, emphatically no, we cannot say when languages diverged, short
of documentary evidence.

If "when" means "these diverged after those" yes, we can say, as
long as "these" are descendents of "those". If they are not descendents,
but "siblings", we cannot say. And I am being accommodating, since
I have argued that reconstructed trees are necessarily unrooted.
But this discussion would take us too far.

Peter T. Daniels

unread,
Oct 13, 2004, 8:36:40 AM10/13/04
to
Andrew Usher wrote:
>
> "Peter T. Daniels" <gram...@worldnet.att.net> wrote in message news:<416BD6...@worldnet.att.net>...
>
> > > > >There doesn't seem to be good reason for treating them differently, as
> > > > >both split off before what we call PIE.
> > > >
> > > > No, Anatolian split off _after_ what I call PIE.
> > >
> > > OK, after your PIE. But the normal reconstructions treat it as if all
> > > languages diverged at the same time,
> >
> > That's the second time you've said that, and it still isn't true. In
> > fact, since you've already been apprised otherwise, it's a lie.
>
> In the other reply you stated that we can't even talk about when
> languages diverged from each other. The above statement would
> contradict that, would it not?

I most certainly did not. Either you're fantasizing, or you're confusing
me with someone else.

> I said, to give a fuller explanation, that the standard reconstruction
> treats all of the traditional subfamilies as significant to the
> structure of PIE.

You didn't say anything of the sort. But are you now suggesting that a
better reconstruction would _not_ treat all the data as significant?

> > > so that PIE will be after the
> > > Hittite split. In fact, it will have some purely Indo-Greek features,
> > > especially if you assume (as was the older practice) that
> > > morphological complexity is only lost.
> > >
> > > > What we know about Etruscan is pitifully little. Etruscan
> > > > can therefore play no significant part in the reconstruction
> > > > of PIE.
> > >
> > > Indeed this is unfortunate; one would really hope that there would be
> > > more, thanks to them having lived with the Romans ffor centuries; but
> > > the Romans don't seem to have been interested in documenting other
> > > people's speech - I don't believe we have anything by them on the
> > > Celtic or Germanic languages.
> > >
> > > Unless this changes, I do not think we will be able to go back to
> > > Proto-Indo-Tyrrhenian (comprehensively, anyway).
> >
> > Indeed, to use such a name is self-delusion.
>
> Where do you think, then, Etruscan came from?

The east somewhere.

Until we can reliably read Etruscan, there's no way of assigning it to a
phylum -- if it is in fact related to any phylum of which evidence
happens to have survived. It needs to be remembered that we only know of
a tiny fraction of all the languages that were spoken in the first (let
alone second and third) millennium BCE.

Jacques Guy

unread,
Oct 14, 2004, 6:29:47 PM10/14/04
to
Andrew Usher wrote:
>
> Jacques Guy <jg...@alphalink.com.au> wrote in message news:<416DB7...@alphalink.com.au>...

>
> > Nowadays I am more inclined
> > to "a bilingual corpus, from which a grammar and a dictionary can be
> > derived". For let's face it: how do you write a description of a
> > language and a dictionary, if not from a corpus of texts
> > ("text" as in "spoken text" in the case of these languages)?
> > The grammar and the dictionary are only
> > an individual linguist's interpretation of the language, extracted from
> > a corpus. Without the evidence--the corpus--I do not think they are
> > worth much. Better than nothing of course, but not awfully much.

> I'm not sure how much difference there is here.


A great deal. An immense difference. In one case you have the
raw data. In the other, someone's interpretation of the data,
with the data missing.

> After all, someone has
> to translate the corpus, and that is just one person's interpretation,
> is it not?

Your informant's, usually.

> And the users of the corpus will not be able to check that,
> as (obviously) they don't speak the language.


They can check if it is internally consistent. They can see if
they can come up with a better analysis of the raw data (aka
corpus).

Thirty years ago, it might have been defensible to argue for
"grammar and dictionary". Those were the days when I had to
lug two Uhers--because one (at least) would break down during
fieldwork (and they're bloody heavy. Eventually, we switched to
El Cheapo National tape recorders. The quality was not as good,
but they did not break down, and batteries lasted a week instead
of a day).

Nowadays there are no excuses. Any PhD thesis on any language
(not extinct language, of course, like Etruscan) should be
accompanied with the complete corpus used and/or gathered by
the researcher. Recording, transcription, the lot. A CD costs
only 30c or so, so there is no excuse anymore for not publishing
one's data.

Peter T. Daniels

unread,
Oct 14, 2004, 7:46:55 AM10/14/04
to
Andrew Usher wrote:
>
> "Douglas G. Kilday" <fuf...@chorus.net> wrote in message news:<416d6...@newspeer2.tds.net>...
>
> <snip all>
>
> Thank you for this analysis. I have only one question: what do you
> mean by 'Archaic Etruscan'? We don't have any inscriptions before 600
> BC, do we?

So you've plunged yourself into this topic without knowing a thing about
Etruscan to start with? Did you try looking even at encyclopedia
articles?

Peter T. Daniels

unread,
Oct 14, 2004, 7:49:54 AM10/14/04
to
Miguel Carrasquer wrote:

>
> On Tue, 12 Oct 2004 13:08:44 GMT, "Peter T. Daniels"
> <gram...@worldnet.att.net> wrote:
>
> >Why is Miguel's interpretation of Etruscan any more valid than anyone
> >else's? How does the scheme you presented (which I snipped) compare with
> >Georgiev's, for instance?
>
> In the first place, it's not _my_ interpretation of
> Etruscan. The grammatical part of the list is largely taken
> from the grammar section of R.S.P. Beekes and L.B. van der
> Meer "De Etrusken spreken" (1991). Un-snipping that:

(Did you say so?)

> PIE Etruscan
>
> * *-m -n (pron.) 'accusative'
> * *-s -s(i) 'genitive'
> *(Hitt.) *-l (pron.) -l(a) 'genitive'
> * *-i -i 'locative/dative'
> * *-nt- -nth 'pres. ptc.'
> * *-to- -tha- 'pf. ptc.'
> * *-kwe -c 'and'
> *(Hitt.) *-ma -m 'and, but'
> * *-bhi -pi 'for, at, ...'
> * *me- (nom.) mi 'I, me'
> * (acc.) mini

> * *k^e- i-ca 'this, that'


> * *to- i-ta 'this, that'
>

> Most of the above is not in any doubt as far as Etruscan is
> concerned: pronouns do have an accusative in -n, nouns and
> pronouns have genitives in -s or -l, locative in -i, etc.

Douglas took exception to nearly every line.

There is _no_ definitive interpretation of Etruscan grammar.

> The existence in Etruscan of a genitive in -l (from *-la) is
> a rare instance where Etruscan *can* shed some light on PIE
> morphology. Hittite pronouns have a genitive in -(V)l, and
> Lydian has an adjectival genitive in -(V)l(i)-, besides the
> usual IE genitives in -(V)s, -(V)s(i)-.

Why isn't that as coincidental as any chance resemblance? If there is
some connection with Etruscan, why isn't it a borrowing, perhaps via
some set phrase?

grapheus

unread,
Oct 14, 2004, 6:59:10 AM10/14/04
to
"Douglas G. Kilday" <fuf...@chorus.net> wrote in message news:<416d6...@newspeer2.tds.net>...
> "Andrew Usher" <k_over...@yahoo.com> wrote ...
> >
> > [...]

> >
> > I did not say that Anatolian and Tyrrhenian were a subgroup, but they
> > they were both related to IE. As far as the Indo-Tyrrhenian
> > hypothesis, I reproduce here Mr. Carrasquer's table of
> > correspondences:

Hereafter my "Kretschmerian comments", which are closer from Mr
Carrasquer's position than Mr Kilday's one :

> >
> > PIE Etruscan
> >
> > * *-m -n (pron.) 'accusative'
>

> Archaic Etr. <-ni> (see <mini> below); Recent Etr. <-n>.

Criticism without value. -n and -ni are two forms of the
"emphasis-particle" -*ne used as a "marker" for the "Kretschmerian
Proto-IE" so-called "accusative" . The primitive meaning of -*m(e/i)
in "Kretschmerian Proto-IE" was the (emphatic) marker meaning "me,
myself".
The confusion between -n and -m is purely "IE stricto sensu". See
J.Faucounau's basic papers on Lycian in the B.S.L. 1982 & 1987.

>
> > * *-s -s(i) 'genitive'
>
> Etr. <-si> is dative.

Valueless objection : TRUE "grammatical cases" do not exist in the
"Kretschmerian Proto-IE" languages. The marker -*s(e/i) marks the
"appartenance, the relationship".

>
> > *(Hitt.) *-l (pron.) -l(a) 'genitive'
>

> Etr. <-la> is found as a genitive only with demonstratives. The Recent
> genitive <-l> replaces Archaic <-ia>; it does not have the same origin as
> <-la>. Nouns taking the non-sibilant genitive also take a non-sibilant
> dative <-(ia)le>.

Valueless objection. There are no TRUE "grammatical cases" in the
"Kretschmerian Proto-IE" languages. The marker -(a)l marks the
"relationship", and has been generalized in Hittite for the
adjectives.

>
> > * *-i -i 'locative/dative'
>
> The function of Etr. <-i> is comitative/instrumental/proximative, like Eng.
> 'by', 'with', despite the tendency of many Etruscanists to lump it together
> with the true locative <-ithi>. Mislabeling <-i> as "locative/dative" goes
> back to Trombetti, who assumed that broad similarity to Anatolian forms
> implied identity.

Incorrect objection . In fact the "Dative-particle" in "Kretschmerian
Proto-IE" is -*ei/i , and became the "true dative-ending" in "IE
stricto sensu". It must not be confused with the "Locative-Marker" -
*thi. Therefore calling it "locative/dative" is misunderstanding the
"Kretschmerian Proto-IE character" of Etruscan.

>
> > * *-nt- -nth 'pres. ptc.'
>

> Not a true participial suffix, but a non-temporal derivative suffix.


>
> > * *-to- -tha- 'pf. ptc.'
>

> False division. The actual (preteritive) participial ending is <-as(a)>.
> This can follow the verbal postfix <-th> as in the participle <svalthas>.
> Falsely regarding this as a genitive leads to extraction of "-tha-".

Pro- and con- arguments are based upon a "non-Kretschmerian" approach.

>
> > * *-kwe -c 'and'
>
> Archaic Etr. <-ka>.

Valueless objection. There is nevertheless a link between "IE stricto
sensu" -kwe and Etruscan -c.
The *ka particle is also at the origin of Greek <kas> and <kai>.

>
> > *(Hitt.) *-ma -m 'and, but'
>

> The sense of Etr. <-m> is sequential, 'and then', like Greek <de>.

Same remark as hereabove for IE -kwe// Etruscan -c.

>
> > * *-bhi -pi 'for, at, ...'
>

> <-pi>/<-pe> has not unequivocally been shown to be a postposition. It could
> just as well be an enclitic honorific, 'Respected One' or the like.
>

I've no opinion on this point.

> > * *me- (nom.) mi 'I, me'
> > * (acc.) mini
>

> Archaic <mini> (late var. <mine>, <mene>), Recent <men>. Many families of
> languages have first-person pronouns in [m-], so this is not definitive
> evidence for tree-building.

See hereabove the § concerning the "Kretschmerian Proto-IE" particle,
marking "emphasis".
The correspondance between "IE stricto sensu" and Etruscan is valid.


>
> > * *ko- i-ca 'this, that'
> > * *to- i-ta 'this, that'
> > * *'ag- ac- 'to drive, to do'
>

> Etr. <ac-> is better rendered 'to bring'.

Doesn't change the link !..

>
> > * *hanti- hanthe 'before'
>
> Trombetti conjectured this meaning _based specifically on resemblance to PIE
> *Hanti-_ whose laryngeal is preserved in Anatolian. In fact <hanthe(c)> is
> an error found once for <hathe(c)> in the Liber Linteus.

I've no established opinion about this point, but I believe that
Kilday is right, this time. There are no laryngeal in Etruscan.

>
> > * *leudh- laut-n 'free(dman), people, family'
>

> The stem of Etr. <lautn>, <lautni>, etc. is borrowed from a pre-Italic IE
> language.

"Ad hoc" explanation to save the DOGMA that "Etruscan has nothing to
do with Indoeuropean" !..

>
> > * *nepot- nefts 'nephew, grandson'
>

> Etr. <nefts> is borrowed from Umbrian. Lemnian <naphoth> is borrowed from
> an unidentified IE language.

Ad hoc explanation !!!

>
> > * *sed- sut(h)-,sat(h)-'to sit, to set'
>

> Etr. <suth-> 'to place' can be extracted from <suthi>, <suthil> 'site;
> tomb'. The sense of <sath-> is unclear. No systematic ablaut-like
> vowel-variation in Etruscan roots can be detected or should be inferred.

The link is nevertheless correct.

>
> > * *dei-n- tin 'day'
> > * *dei-u- tiu(r) 'light > moon, month'
>

> <Tin> is most commonly 'Zeus' and both of these words are best explained as
> religious borrowings from pre-Italic Indo-European.

Ad hoc explanation to save the DOGMA !...

>
> > * *doh- tur 'to give' (cf. Grk. do:-r-on)
>

> Another religious borrowing from pre-Italic IE. <tur-> is found only in
> votive contexts.

NEW AD HOC explanation to save the DOGMA !..

> The secular Etruscan word is <al-> 'to give'.

Dubious statement. <al-> means "the sky", <als'i > :"the day".

>
> > * *dem-/dom- tmia 'house, temple'
>

> Etr. <tmia> is attested only in the Pyrgi temple-dedication, where it
> clearly refers to the temple. Yet another religious borrowing from IE.

Dubious statement. No proof.

>
> > * tham- 'to settle, to establish'
>

> What is the Etruscan verb being compared with?
>
> > * *sahu-l- usil 'sun'
>
> Etr. <usil> is either native or borrowed from a word related to Sabine
> *<ausel> which can be inferred from the name <Aurelius>. One explanation of
> *<ausel> is taboo scrambling of IE *sawel-, but that does not make <usil>
> evidence for "Indo-Tyrrhenian".

Here, I share Kilday's opinion. In the Pyrgi inscription, the
"Sun-God" is <ila(cve)>.

>
> After borrowings and errors are discounted, all that is left are a very few
> similarities which do not even follow a consistent set of sound-laws. This
> is not enough to found a credible theory upon.

Better said that this "Approach to Etruscan" of both Mr Carrasquer and
Mr Kilday is viciated by the failure to recognize the "Kretschmerian
Proto-IE" character of Etruscan !..

grapheus

Peter T. Daniels

unread,
Oct 14, 2004, 7:45:57 AM10/14/04