SV: Iti & ti

<Thus, like K.R. Norman's example of "Brahma" vs. "Bahama" we may have a
case of a deleted Prakritic element of the language in the reduplication of
"t"
in "itti">

The example of Pali brah- vs Ashokan bah- is often adduced as an example of
a Sanskritism on the assumption that Pali necessarily must reflect other
similar or historically close Prakrit(s). I beg to disagree. A look into a
Pali dictionary shows that Pali does not disallow the cluster /br/ (e.g.
bruusi, abravi etc.), whereas the language disallows the cluster /pr/. The
evidence is overwhelming and the distribution systematic. On the other hand,
Pali disallows the cluster /bhr/(e.g. bhuunahu/ha[cf. Sanskrit
bhruu.nagha]). In the first case the sonant labial /b/ and the liquid
continuant /r/ are not phonetically exclusive in contrast to surd /p/ and
/r/. The reason why Pali disallows /bhr/ is undoubtedly because aspiration
and /r/ are mutually exclusive. Those who assume that the early compilers
and later generations introduced Sanskritisms like brah-/braah- should
explain why /r/ is regularly elided from the cluster /pr/ but not from /br/.
Clearly the onus is on them.

Regards,
OP