I'm just remembering that I have in my possession another book written by Geoffrey Hull: Polyglot Italy.
He dedicates a chapter to the "dialects of Sicily, Calabria and Salento".
"...the Italian dialects of these regions are not local developments of Vulgar Latin but modern forms of a southern variety of Italian introduced not more than a thousand years ago into what used to be Greek and Arabic-speaking territory.
...
The Byzantine invasion of 535 AD enabled Greek to triumph over Latin throughout Sicily, at least on an official level, but it is likely that vernacular Romance lingered on the west of the island...
By the early eleventh century these islands and the whole of Sicily except the Hellenophone north-east had become predominantly Arabic in language...
There seems to be no conclusive evidence that any form of Romance speech survived the Saracen occupation of Sicily (or that it had even been vital there during the Byzantine period). Certainly the Italian speech that came to Sicily in the wake of the Norman conquest (1060-1091) was quite new to the island.
...Romance dialects of the two main groups of settlers brought in by the Hautevilles to reduce the Arab and Greek presence in Sicily. The Campanians (peasants from the Naples district and possibly from Cilento as well) colonised the Arabicized territory west of the Salso.
...
Most of the settlements in the east of the island were founded by pioneers from the Monferrat region of the western Po Valley. Padanian immigration to Sicily had been promoted by Roger I's marriage to Adelaide, daughter of the Marquis of Monferrat..and her brother Henry, married to Roger I's daughter Blandina, ruled the County of Paterno, the nucleus of Lombard settlement in eastern Sicily.
Norman Sicilian was slower in displacing the Padanian speech of these newcomers, but before its extinction in most areas this Northern language injected a substantial body of Gallo-Romance elements into the koine...the common Sicilian vocabulary had long since absorbed such fundamental Padanian words as soggiru (suoxer), cugnatu (cognau), bizzuni (besson), figghiozzu (figlioz), fadali (faudal), tuma, tumazzu (toma) fugazza (fogaza), orbu (orb), arricintari (rexentar), unni (ond), and the names of the week: luni (lunes), marti (martes), mercuri (mercor), jovi (juovia), venniri (venner).
Padanian influences on Sicilian extend into the realms of morphology and syntax...
The double Gallo-Romance (Old French and Padanian) impact on Sicilian accounts to a large extent for the relatively 'modern' (ie non-archaic) character of the dialect in comparison with vernaculars like Sardinian and Lucanian, the results of an uninterrupted bimillennial tradition of Latinity.
As Italo-Greek declined in neighboring Calabria it was replaced by the Messinese variety of Sicilian.
The neo-Italian (i.e. post-hellenic) speech of the Salentine penninsula (Terra d'Otranto) is not a form of Sicilian like that of lower Calabria, but represents the medieval Apulian dialect that filtered into the district during the period of civil reconstruction after the devastiting Moorish raids of the tenth century and the Norman conquest of 1054-1068. If Salentine appears superficially very similar to Sicilian it is because the Old Apulian from which it derives was itself not far removed from the South Italian koine that formed the basis of Norman Sicilian. The dialect has been strongly influenced by the local substratal Greek, but naturally lacks the Arabic and Padanian elements so prominent in the language of Sicily.
Professor Alberto Varvaro: " It is well known that opinion varies greatly on the linguistic situation that the Normans found in Sicily: no-one any longer doubts that a large part of the population, the administration and cultural milieux used Arabic, or is the survival of Greek (at least in the north-eastern area) a matter of controversy. Rather it is the existence and numbers of people speaking a neo-latin language that is uncertain and open to question.