'Sandhi' refers to changes in pronunciation which take place when words or pieces of
words (that is, stems and affixes) are combined. I would assume that the spellings
'sa"ngha' and 'sa.mgha' do not represent any difference in pronunciation. If that is the
case, then the spelling cannot be itself an instance of sandhi. It is, however, a result of
sandhi in the following sense: the word in question contains the prefix 'sa.m-' which
means something like 'together'. It is found in many words where the spelling .m would
have a different pronunciation than "n; the word 'sandhi' itself is an example, and you will
sometimes see it written as 'sa.mdhi'. The rule which Jim refers to specifies that before a
velar consonant (k, kh, g, gh) .m may be pronounced as a velar nasal ("n). So we might
say that the existence of the rule allows the phonetic details of "n to be dispensed with in
the writing system. It is interesting that in Sanskrit as well as Pali all cases of "n (outside
phonetic treatises and grammars) result from this sandhi rule, so that the letter "n itself is
dispensable.

--- In Pali@yahoogroups.com, "Jim Anderson" <jimanderson.on@...> wrote:
>
> Dear Florent,
>
> You're right, the two rules (Kacc 30 & 31) concern euphonic
> combinations. And, yes, the two words are a result of sandhi although
> I find some difficulty understanding some of the commentarial remarks
> on Kacc 30 (a.m bya~njane niggahita.m) which I think applies to
> "sa.mgha". With Kacc 31 (vagganta.m vaa vagge) it is easier to see
> that .m is changed to "n before gh in the case of "sa"ngha". It is
> possible that "vaa" may not have the meaning that I've been thinking
> it has.
>
> Jim
>
> > Dear Jim,
> >
> > Thanks a lot for your explanation. However I thought this rule was
> > concerning euphonic combinations mainly (sandhi). In this case can
> we
> > also say that the word "sa.mgha" or "sa"ngha" is the result of a
> Sandhi?
> >
> > Kind regards,
> >
> > Florent
>