Three-Way Contrast of Secondary Articulations in PIE

[Stanley Friesen]

At 05:02 PM 5/7/01 -0500, proto-language wrote:
>Dear Stanley and IEists:
>[PCR]
>The normal formation is from *bhu-n-g-ti or *bhu-n-kti. Yes, *bhunegti can be
>reconstructed on the basis Old Indian bhunákti but this is a fish swimming
>against the stream.

I was following the traditional reconstruction.  I cannot at this time
evaluate the relative likelihood of the various alternatives.

>> While I agree many of the roots probably originally were distinct, I do not
>> think we yet have sufficient information to tell in what manner.  I
>> certainly doubt there was a single cause for all of the mergers.

>[PCR]
>First, our agreement: there is rarely a single cause for anything.

>But -- putting glides aside, how were they kept separate?

My point is that in PIE per se, they weren't, except by formatives such as
the noun stem formatives.  That is, by the time of the reconstructed
language, the old conditioning factors were gone.

>Glides is the most parsimonious explanation.

I would tend to say, we do not know what the differentiating factors were
in the pre-stage preceding the reconstructed stage.

Also, I suspect that multiple factors kept them separate at that stage.

--------------
May the peace of God be with you.         sarima at friesen.net