DDeden Wrote:
-------------------------------------------------------

> Yes, very interesting, words with dissimilar roots
> can appear to be derived from the same roots.
> But then, the opposite occurs as well. For
> example, 2 in Italian (IIRC) is 'dua', and 2 in
> Indonesian is 'dua', and I've read that they did
> not derive from the same root, but I that they
> did.

But the Indonesian word does not exist in a vacuum. There're related words in the vast Austronesian (AN) family to which Indonesian belongs. Underlying all of them historically is Proto-Austronesian "duSa", with a sibilant that was lost in some (but not all) AN languages. As you can see, the OLD form is LESS similar to the word 'two' in Indo-European (IE) languages (they go back to animate *dwo:, inanimate *dwoi(h), with endings representing the IE dual number). That's usually the case with spurious cognates. They look similar when you compare modern languages but the similarity evaporates as you move back in time. True cognates behave differently. They may not be similar at all at present, but as you trace back their history, the similarity INCREASES.

In this particular case the similarity is isolated. It isn'r part of a more general pattern of regular correspondences. This means that you cannot rule out the possibility that it's a chance resemblance (and that's what it most likely is).

> I asked about ak- aht- ax- in relation to
> hammering or chopping, as I think it was conjoined
> with watr to make akwa (dugout canoe, later other
> related words). I think 'umiak', 'kayak' (names
> for ribbed skin boats), 'Ottawa' derived partly
> from the Altaic (Ket) word for watercraft,

Pleeeeeeez. "Ottawa" comes from the Ojibwe tribal name "Odaawaa", meaning 'the Traders'.

> and I
> think similar derived words would be found in
> numerous languages not closely related. A couple
> of AE words for watercraft include these sounds,
> and I'm wondering if akwa referred only to boats
> derived from dugouts or if the name for reed raft
> boats was also derived from it, since they were
> constructed entirely differently.

You're repeating the same error again. The fact that words for similar things in different langauges seem to contain the same sounds is no proof of anything. These "similarities" are indistinguishable from what is produced by pure chance.

> Your PIE etc. language site is great, I enjoyed
> reading it.

Thank you. Please, note in particular the cautionary examples of false cognates discussed there.

Piotr Gasiorowski