Re: 'Dog' revisited

> >
> > > >
> > > > What exactly _is_ cherry-picking, and why is it inadmissible?
> ................
> > The way I was using it was to mean picking out similar sounding
> > terms
> > > and not accounting for there overall frequency.
> ......
> >
> > No one uses statistics that way within a recognised language

group.

>
> Since the mass comparison method is used for the purpose of
> discerning supposed groups it is not relevant for already

recognized

> groups. Hence no need for such statistical counter-arguments.

That is not the way I use "mass comparison". I tend to scavenge
already published attempts at proving new super-groups using mass
comparison for words I then claim to be loans.

The reason I do that is this: Whenever you try to do mass comparison
you find that most words are thoroughly incompatible with the
exception of a few which are suspiciously lucidly similar. I'm
reminded of the experience I get listening to a Greenlandic newscast
on the radio, sounding in my ears like this "groqpoq kommunaldirektør
voqgioq syvogtredive millionit kronit toq"; in other words I suspect
they are loans. Why else is Ruhlen's list so short? Obviously, if you
try other words, you fail to come up with correspondences.

>
> > Ten years ago your argument might have been used against

Nostratic,

> > now it isn't, which is another way of calling it ephemeral.
>
> Perhaps it is only another way of saying the critics of the

Nostratic

> hypothesis have seen nothing in the last ten years which they
> consider new arguments worthy of a fresh look. They have moved on

to

> other things, and see no point in continually reiterating past
> criticisms.
>

Do not despair! There are always new things somewhere to be
criticised ;-)

> > Personally, I've mostly left out New World cognates, since I know
> > next to nothing about their development. Other than that I've

used

> > Møller's IE-Semitic and Bomhard's Nostratic etymologies (on the
> > assumption that some of them are loans) plus Austronesian (not

all

> > proto-, true) and Proto-Bantu. You don't get back much further

than

> > that.
>
> So that if one accepts the Nostratic hypothesis, then we are

dealing

> only with a case of three language groups happening to have a

similar

> sounding term for dog (assuming your Austronesian examples
> representing the original form)?

Whether or not one accepts the Nostratic theory, then, given the
AfroAsiatic evidence, that *kwon- word was not part of it.

>How do we rule out coincidence? And
> is the inclusion of proto-Bantu another instance of (forgive the
> word) cherry-picking?

Actually I've used several other correspondences with Proto-Bantu.
The reason I included it was this: If there is an Austronesian
influence on IE it must have gone either overland in the Middle East
or around Africa. Logistically it would be the reverse of the
European finding a trade route to the East. In order to create a
trade route overland you need control over a route to the extent that
you are not disturbed by tax-happy local sovereigns. The Europeans
tried for centuries to break through the Muslim states to create a
safe passage to the Indian Ocean but it wasn't logistically possible.
Only with da Gama's journey did they find a way. And this alternative
route financially weakened the Middle East to a degree that it then
became possible to create an overland route.
In order to have a circum-Africa route you need way-stations. Paul
Manansala refers to a people (the name of which escapes me) who
sailed between the Philippines and China and settled deltas of
rivers, and later went further up the rivers, and brought new
civlisations to the ares they settled. The obvious choice in West
Africa is the Niger delta: they grow the third agricultural species
of rice (the two others are Chinese and Indian), the Bantu expansion
started there, they have iron (Proto-Bantu *-beda (Meeusen), PIE
*bherso-m > Latin ferrum, Semitic b-r-z, Proto-HesperonesianFormosan
*bari[], Proto-Hesperonesian *besi), and in general, for whatever it
seems many of the words on Manansala's and my list seem to have
cognates in Proto-Bantu (at least one rule: l/r > d).

>What is the criteria for including them rather
> than another group in the comparison? Is it just because they do
> happen to have a similar sounding term?

See above.

>
> >
> > > > >
> > > > > I see now- a conjectural etymology
> > > >
> > > > As opposed to what kind of etymology?
> > >
> > > As opposed to a generally accepted etymology
> >
> > Which all began as conjectural etymologies.
>
> Which is besides the point- the point being that many, if not most
> conjectural etymologies do not become generally accepted ones.
>

Aha, and you don't accept them, so they are not genrally accepted ;-)

> > > ....... I asked if
> > there
> > > was evidence of voyages to Taiwan before the Neolithic?
> >
> > At the time of low water, Taiwan was highland, relatively. Why
> should
> > the inhabitants of the river plain go there? The way I see it,

the

> > Austronesian speakers of Taiwan are refugees from the floods.
>

> Is this a way of conceding that you do not have evidence of
> paleolithic canoe voyages to Taiwan?

Yes. On the other hand we don't have evidence of canoe voyages to the
Mt Blanc either.

>I had asked because you were
> implying the proto-Austronesians gained seafaring experience by
> making trips across the widening Taiwan Straits. Are you now saying
> that maybe they didn't make those trips?
>

No. They probably moved up river. And then there were plenty of new
straits opening for them to practise on.

> > > .... But if for the sake of discussion I were to
> > > concede your point, then my question would be what would your
> > > hypothesis explain which could not be explained by Proto-World?
> >
> > Proto-World would have been much older than the domestication of
> > dogs. The first emigration out of Africa followed the coast of
> > Southern Asia. Dogs being domesticated in East Asia would have to
> > have gone the opposite way. In other words, for a *k-n-, *k-l-
> > etc "dog" word to be Proto-World, dogs would have had to be
> > domesticated for the first time in Africa.
>
> Yes, I was aware of that point when I asked the question- but there
> could have been a term for wild dogs, which could have been used

for

> domesticated ones too when they appeared on the scene.

But I'm wary of the reconstructibilty of Proto-World for the reasons
I gave above.

> >
> >
> > >I'm
> > > certainly no devotee of Proto-World, but if confronted with a
> > choice
> > > of just that or the "transcontinental canine trading

paleolithic

> > > proto-austronesian canoe paddler" hypothesis I am not sure

which

> > one
> > > Brother Oakham would start shaving.
> > >
> >
> > I haven't proposed these people bartered dogs.
>
> Well, actually you did. As you may recall, my initial entry into

this

> thread was prompted by your description of dogs as trade items
> carried by the Austronesians.
>

I am sorry if I might have upset some British sensibilities ;-) I
never intended to give the impression that those traders traded their
dogs. Besides, it would simply be uneconomical.

> > I think they brought
> > them with them. Maybe they gave away a puppy or two to the

natives?

>
> OK, I don't want to beat the dead horse re trade here- we both
> concede that dogs could have been traded at times, and could have
> spread by other means at other times. But we remain divided over
> whether there is evidence to support a supposition that

Austronesians

> played the major role in their spread, and in the spread of their

name

> [s].
>

True.

Torsten