Classical Latin appears to be without regional dialects, yet Latin evolved in little more than a millennium into a variety of different languages (the Romance Italian, French, Spanish, Portuguese etc.). Was regional diversity apparent from the earliest times, obscured perhaps by the standardisation of writing, or did some catastrophic event in late antiquity cause the language to vary? These questions have long intrigued Latinists and Romance philologists, struck by the apparent uniformity of Latin alongside the variety of Romance. This book, first published in 2007, establishes that Latin was never geographically uniform. The changing patterns of diversity and the determinants of variation are examined from the time of the early inscriptions of Italy, through to late antiquity and the beginnings of the Romance dialects in the western Roman provinces. This is the most comprehensive treatment ever undertaken of the regional diversification of Latin throughout its history in the Roman period.
Tediously philological and to some degree self-sabotagingly synchronic—Adams methodically works from texts and epigraphy and only incidentally looks at modern reflexes, but it's when he does that the book is at its best, despite his expressed ambivalence about that as a method. The front cover promises maps; there are eleven of those in an appendix, and while they're pretty lovely (they're all in dire need of dating), they build unrealistic expectations of how well-structured and to-the-point the rest of the book is going to be. That veers from (occasionally) laborious and barely coherent to (more often, especially towards the end) very compelling, but even when it's good Adams' love of running text makes it much less useful as a reference than it could have been—sometimes barely more useful than doing the work yourself.
The general tenor of Adams' conclusions is always that we know less than has often been claimed we know, and that much previous work was fundamentally wrong-headed, sometimes to the point of incompetence. This is often the case for alleged lexical regionalisms, but especially for grammatical and phonological regionalisms, and much time is spent criticising specific academics and techniques. More rigour is always good, but the repeated harping on methodological points does tend to overwhelm the positive works Adams does: in his final conclusion at the very end of the book he's clearly convinced (non-lexical) regional variation in Latin has been demonstrated and the unitarian position is definitively debunked, but that is not what I had been able to extract from the hemming and hawing. What he did show fairly convincingly—and what he also highlights in that conclusion—is that the apparently common view that the imperial provinces were repositories for archaisms collapses under more systematic scrutiny, with purported archaisms typically not being archaic, being deliberately affected style, or even being innovations. (Whether his characterisation of this view as flatly that the Latin of a province reflects that of Rome at the time of its annexation ("The language of Sicily reflects the Latin of Plautus, Spanish that of Ennius, French that of Caesar, and so on.") is fair is less clear.)
Ultimately I'd say a lot of work obviously went into this but it didn't pay off quite as well as it might have. Still useful; not worth reading cover to cover like I did.
Although Latin was spoken over a huge area, from North Africa to Britain, from Spain to the Balkans, the texts that have come down to us present a remarkably homogeneous language. There must have been differences between one region and another in pronunciation and lexis – eventually we got all the various Romance languages from them – but past attempts to identify these differences have been fraught with problems. This massive work by J. N. Adams lays out what was wrong with earlier attempts and, working on a stronger methodological basis, tries to make some more firmly grounded claims for distinct regionalisms in various parts of the Roman Empire.
Various Roman authors made claims that a certain feature was a marker of a certain region, but Adams cautions against accepting these claims uncritically. For example, during 20th-century surveys of British dialects, many speakers believed that a word was something special that only their area had, but it was in fact widespread across Britain and those speakers just weren't informed about other regional varieties. Another issue is Roman authorities' inability to distinguish between regional varieties and sociolinguistic registers, whereby a feature they claimed was "rustic" was likely found even within Rome itself, just among lower social classes than their own.
Not only must Roman authors be taken with a grain of salt, but 19th and 20th-century scholars make a number of grave mistakes. In pointing to inscriptions as a sign of regional features, they failed to take into account that engravers may have perpetuated archaic spellings that didn't accurately reflect the local language. Adams also makes a number of devastating critiques of the claim, especially popular in Spain, that regional Latin varieties must have preserved Latin as it was spoken at the the time that the particular region was conquered.
Adams' new, refined case for identifying regional varieties is divided into two parts, the first examining evidence from the Roman Republic, and the second from the Roman Empire. This reflects the changing sociolinguistics of Latin, where initially Roman authors thought of themselves versus the Italian countryside, but with the Empire the opposition was more broadly between Italy and the rest of the Empire. He exhaustively looks at virtually every single mention of regionalisms found in Roman authors, although many must be discounted due to the flaws mentioned above. He also cautiously looks at inscriptions with certain caveats, and he also makes use of the curse tablets and Vindolanda documents that have come to light in recent decades.
One should not expect any kind of detailed description of individual regional varieties. Ultimately, Adams offers only a relative handful of regional terms and vowel shifts, and these are hard won. The Regional Diversification of Latin tears down the shaky edifice of two centuries of ungrounded claims by classical philologists, but the solid facts that can be built on the rubble are fewer than one would have wished. Still, as a work of scholarship this is an awesome achievement, and this linguist can only marvel at Adams' dedication and breadth of knowledge.
The reference for the various regional changes Vulgar Latin experienced, Adams covers the entire Empire and all brands of Latin, from that of Gaul to British and African Latin and their particularities.