Hittite & Celtic dative in /k/ ?

[ECOLING at aol.com]

In a message dated 8/3/99 6:49:38 PM, 114064.1241 at compuserve.com writes:

>the comparison is unlikely, as the Hittite form is also used for the
>accusative, an use for which it has cognates in Germanic (gotic mik) and
>perhaps in Slavic (russian ko - to with only an allative meaning). The
>primary meaning was probably allative, as for most accusative in IE.

This kind of information in no way lessens the probability of connection,
since (as stated just above) the accusative may derive from an older
dative (or allative or etc.).  This kind of information CAN suggest that
a connection may have great(er) time depth, to allow time (how much?) for
a common change to occur in one or another language.
(I assume it is not being proposed that the ORIGINAL meaning was
accusative, and that the accusative developed into a dative or allative?
That would be a case of a more grammatical or "bleached" meaning changing
into a more concrete meaning, possible of course but less likely.)

As a modern example, consider the use of the "dative" preposition "a"
(if one wishes to call it that, where we do not have inflectional cases)
with Spanish animate direct objects.

>the Goidelic form is assumed to derive from *angh "near"
>which is found in Latin angustus.
>Its closest cognates, assuming to Stokes, is Breton hag
>(and) and Welsh ac (same meaning)

I am not as familiar with typological tendencies for an
adessive (being "at" or "near") to develop either into a dative
or into an accusative.

Whether the speculated connection is valid or not is quite another question.

I just do not see how the offered information (except the *angh part)
has a bearing other than to strengthen the plausibility.

Until someone finds another example of a postposition "near"
turning into a dative/accusative etc., in some language of the world,
I am doubtful of this etymology
on typological grounds.
I would love to have my typological horizons expanded if this is really a
clear case,
with intermediate stages actually attested.

Lloyd Anderson