>Perniola writes that the adjective 'matta' ("by measure, measured")
>"is often used at the end of a bahubbiihi samaasa in the sense of
>'in the measure of, as much aas, as little as, merely, only.'" He
>gives several examples (171), including: maran.na-matta.m dukkha.m,
>"a pain as great as death."

Sure, but Perniola (and the PED) are simplifying the matter for
practical reasons (and confusing it along the way). There is no
adjective 'matta' used at the end of bahubbiihis. There is a noun
mattaa (fem), which when place at the end of a bahubbiihi cpd turns
into an -a stem form (like all bahubbiihis) and which literally means
'Y has X as it's measure': X-matta Y.

Since this is so common, one can teach students to watch out for the
'adjective' -matta ifc meaning 'by measure' but this is just a
pedagogical tool. Or if you choose to view this as no longer a result
of a productive bahubbiihi formation and instead a newly naturalized
adjectival form, then the result is no longer a bahubbiihi at all,
but just a normal adjectival kammadhaaraya compound.

If a noun turns into an adjective through being included in a
bahubbiihi, you can't then go and call the resulting form a case of
an adjective concluding a bahubbiihi. You're conflating how the form
was composed, with the final product.

best regards,

/Rett