Ix-nay on the ostratic-nay

[Miguel Carrasquer Vidal]

iffr762 at utxvms.cc.utexas.edu wrote:

>	With regard to reconstructing 1sg pronouns, there is (or so I seem
>to recall) a cross-linguistic tendency for these to be formed with /m-/,
>probably from a weaker variant of what might be called the "mama
>syndrome": sounds that babies tend to make early tend to be pressed into
>service as words that mamas and babies might use to relate to each other,
>like "mama" and "me".

I don't think so.

>	So seeing a 1sg in /m-/ does not necessarily mean much.

This is true (although seeing 1sg in /m-/ AND 2sg. in /t-/, or
1sg. in /n-/ AND 2sg. in /k-/ may definitely mean more).

There is, I think, a cross-linguistic tendency for (personal)
pronouns to be short forms.  For instance, in PIE they show a
pattern CV where verbal and nominal roots have at least CVC.  The
same thing is true for many other language families, although I
have not made a study of this.  In that sense, seeing a 1sg. form
apparently based on *mV- doesn't mean much in itself, especially
given the fact that /m/ is a phoneme that is present and common
in the overwhelming majority of the world's languages.  I also
haven't made a study of phoneme frequencies the world over, but
my impression is that the general monosyllabicity of pronominal
stems and the relatively high frequency of the phoneme /m/, plus
"number of pronominal stems to be distinguished" versus "number
of phonemes in the language" are sufficient to explain the number
of languages that have 1p.sg. *mV- (which really isn't *that*
large, I'd say 10-20%).  No need for a "mama-effect".

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl
Amsterdam