c/p/t behind n/m/r/l less likely to become g/b/d in Romance?

[Lusus Naturae]

Many of them remain surd, like alto, barca, puerta, sempre, viento, ...
But I've seen conde < comitem and linde < limitem, are there many more examples of c/p/t behind n/m/r/l becoming g/b/d?

 

[ahvalj]

As far as I understand, c/p/t after n/m/r/l don't become voiced in Romance (at least not in the dialects at the basis of the standard languages). The cases with voicing, as can be seen from your examples, have it occurred before the syncope, i. e. comitem>*comide>conde, also aliquod>*aligo>algo, *debita>*debida>debda>deuda.

 

[Riverplatense]

[A]re there many more examples of c/p/t behind n/m/r/l becoming g/b/d?

In some Southern Italian dialects [p t k] are voiced after nasals: MONTEM > mondë.

 

[Zec]

And in French it's unpredictable whether syncope will happen before voicing, or voicing before syncope.

 

[Penyafort]

There are several instances of such a voicing in Aragonese (aldo, cambo < CAMPU, chungo < JUNCU, etc) which are even more widespread in toponymy (Pandicosa, Benduso < VENTOSU, etc)

 

[Lusus Naturae]

I saw an opposite example: gangraena > cancrena.
How rare are they?

 

[Riverplatense]

I saw an opposite example: gangraena > cancrena.
How rare are they?

This case is not actually a sound shift based on articulatory phenomena, but an adopted popular etymology: Cancrena o ganmena. I'd suppose spontaneous desonorisation between nasal and liquid to be very rare.

 

[danielstan]

Another example of popular etymology is the contamination of Latin crassus with grossus which gave in some Romance languages: gras.

crassus - Wiktionary