Re: [tied] -ch

>> In Czech, expressive forms of family relations and personal names
>> are derived with the suffix -ch [x] added to first syllable of the word,
>> for example:
>> bratr > brach, kmotr > kmoch,

another appellative with this suffix seems to be
hoch "boy" < hol-ec "the hairless (thus unmature) one"
(fem. hol-ka "girl).

>> Petr > Pech, Matej > Mach ...
>
> It was the same in Old Polish
> There are also related East Slavic hypocoristics (cf. -s^a, -s^ka)

It is interesting that such -a forms seem to be sometimes from -ja
and sometimes from -a only (-s^a < -x-ja, -s^ka < -x-ik-a).
In Czech, the suffix [-x] before this -a suffix is not palatalised:
brácha (bratr, brach, brás^ka), Hácha (Havel), Mácha (Mate^j, Mach).

Unlike the [-x] suffix, which is added to words shortened to open syllable,
the -ja suffix is added usually to words shortened to closed syllable:
Petr > Pe-ch but Pét-ja,
but there are exceptions as Karel > Kája.

>> My guess is that it is augmentative suffix -sko
>> with expressive metathesis to -kso
>> and RUKI change to [-xo].
>
> This looks completely ad hoc to me, though I can't offer a convincing
> alternative off the top of my head.

I was not aware of any other source of [x] than RUKI,
but now I found a mention that [x] can be also from PIE. k^H2.

P.A.