For a while now I've been playing with the palatised c sound. I've been pronouncing it actually palatised, with the blade of my tongue against the palate rather than the tip against the alveolar ridge. Logically one can imagine the velar consonant travelling to the alveolar ridge by way of the palate. The fun thing with this pronunciation is that this particular consonant doesn't exist in modern English and as a result if you're listening for ch you hear it, but if you're listening for velar c, you hear that too. This pronunciation makes sense makes sense, but then again language rarely makes sense. The c could just as easily have flipped from a velar to an alveolar consonant.
In terms of borrowing. If the initial borrowing happening after the active sound change then the original sounds would have been preserved. We look at Cerdic and modern English spelling rules suggest the pronunciation Serdick. Pre-conquest spelling rules suggest Chair-ditch. However, these guys weren't reading the name. If you look at something like the drink curacao. Those that have never heard the Portuguese pronounce it cuRAcow. Those that have pronounce it CUraSOW (capitals to show word stress).
In conclusion, we don't know how it was pronounced, but Chair-ditch is probable, although it's certainly fun to muse upon these things.
