I'd like to know how the pronunciation of the letter 'C' has developed in Latin. All I know so far is that it has changed through the centuries, but I'm interested in specifically what those changes were. I was wondering what others here might know regarding this development.
For context, none of the pronunciation systems used today for Latin developed by continuous change from the pronunciation of Latin as it was originally spoken by the Romans. The natural sound changes that occurred for Latin speakers resulted instead in the Romance languages.
Present-day pronunciations of Latin are all based on spelling pronunciation, or more recently, on scholarly reconstructions of what Classical Latin is supposed to have sounded like.
The original pronunciation of Latin C is reconstructed as [k], a voiceless velar plosive.
Before a front vowel, this may have had a somewhat palatalized/fronted pronunciation (see the following question: (Why do I find it hard not to palatalize the /g/ in digitus?)), but not to the extent of causing it to become an affricate as in modern Italian città. For an example of what a palatalized/fronted variant of Classical Latin /k/ would have sounded like, compare the consonant at the start of modern French qui, which some phoneticians transcribe as a palatal plosive [c], but which is treated in the sound system of French as the phoneme /k/.
Changes between Latin and the Romance languages created other sounds.
Postclassicially, the presence of palatalization caused /k/ to turn into an affricate in some positions, and then develop further to other sounds in some Romance languages. Those developments are multi-part and differ between languages.
The earliest context where /k/ developed into an affricate was probably in /kj/ sequences, which arose from Classical Latin /ki/ or /ke/ before a vowel. An affricate or fricative outcome is found in all Romance languages that I am familiar with, including Sardinian. I think Latin lancea to Italian lancia, French lance, Sardinian lantza is an example. Between vowels, the outcome is identical to that of /kk/ before /j/: e.g. from Latin faciēs we have Italian faccia, French face, from Latin nūtrīcia French nourrice.
The outcome of /k/ before /j/ often is kept distinct from the outcome of /t/ before /j/, even though /t/ also developed an affricated pronunciation in this context at some point.
In /ki/ and /ke/ not followed by a vowel, an affricate or fricative developed in Italo-Western romance (Italian /tʃ/, French /s/, Castilian Spanish /θ/) and in Romanian, but Sardinian retained a plosive pronunciation. This suggests that affricates developed later in this position than before /j/.
There is a separate question about the timing of this development: When did “c” before “e” or “i” start to be pronounced as [ts] (in contrast to classical [k])?
Those other sounds became used as spelling pronunciations in the pronunciation of Latin by Romance speakers
After Latin had become established as a distinct entity from the regional varieties of Romance, it became usual to pronounce c before the letters i (y) and e (ae/oe) with the sound that each of these languages had come to use for C before these letters. E.g. /tʃ/ in Italian, /s/ in modern French, /θ/ in modern Castilian Spanish. In the Roman Catholic Church, it came to be thought preferable for members in all areas to follow the pronunciation used in Rome, so /tʃ/ became established as what is often called the "Ecclesiastical Latin" pronunciation of C before E/AE/OE/I/Y.