Amid the annual deluge of miscellaneous edited volumes of no special distinction, there every so often arrives a rarity like this one: a coherent collection of essays worth reading through from end to end. Its eight topical chapters, flanked by substantive introduction and conclusion, provide a fairly comprehensive overview of linguistic change and survival in Africa, Spain, Gaul and Germany, and Britain and Ireland, though unfortunately not Italy or the Alpine provinces, from the fourth to the seventh/eighth centuries. There are contributions from both linguists and historians, which is one of the book's great virtues. Historians can benefit enormously from understanding the language changes going on below the surface of their narrative evidence, but monographic linguistic studies tend to be such hard going that non-linguists avoid them. What is more, old-fashioned historical philology, even when dressed up as palaeolinguistics, could at times reach bizarre conclusions because of its dislocation from the fixed points of historical chronology (one thinks of D. H. Green's 1998 Language and History in the Early Germanic World). The present volume will remind, or perhaps inform, historians of Late Antiquity and the early middle ages that modern sociolinguistics is much more firmly rooted in social and historical context than were some of its disciplinary ancestors; that it carries none of the modern political baggage of germanisches Altertumskunde that caused the post-War generation and its scholarly progeny to shun Germanic philology more generally; and that language, among the many things that changed dramatically during Late and post-Roman Antiquity, can reveal social changes that are invisible in other sorts of evidence.

Several themes run throughout the volume. One is the impact of power relations on the use of one or another available language, both in terms of the relative political power of different linguistic populations (Latin versus Punic, Anglo-Saxon versus Romano-British) but, more subtly, in the way language choices encode strategies about how to project power: for instance the Dyfed Stone (29–30, 228–9), with ogam on the edges spelling out the name Votecorigas (a compound of British and Irish Celtic) and Latin on the face reading memoria Voteporigis Protictoris, in which the onomastic elements of the Celtic name could be translated as the Latin protector, a word that may still have recalled the privileged corps of imperial protectores domestici. Placed by the roadside, as it was, the stone might very well have conveyed, to speakers of three different languages, the same message about the man's local prominence. A second, if unsurprising, theme is linguistic complexity — simple paradigms of replacement (one-in, one-out, the editors call it) will not do, and where the evidentiary base is most robust, it reveals just how localised linguistic change can be. On that front, Wolfgang Haubrichs's ‘Death and survival of Latin in the empire west of the Rhine’ is a special tour de force. The focus here is on the Belgic and German provinces of the empire between Rhine and Somme, centred on what would become Carolingian Lotharingia where the Romance-Germanic boundary lies to this day. Because the area is rich in toponyms with close chronological attestations, it possesses a uniquely valuable evidentiary corpus for showing stages by which place-names made the transition from their lexical origins to the adoption of new features from the target language, along with any intermediate stages on that route. What we find are pockets of Latin continuity throughout the region, for instance in the Mosel corridor from Koblenz up to Metz, or around Cologne and Strasbourg, but also places where Frankish was adopted earlier than one might have expected, as on the Meuse near Verdun.

Haubrichs's paper is complemented by two others on developments in the Gallic provinces, one of them Alderik Blom's study of the survival (actually non-survival) of Gaulish, the other Ian Wood on the familiar literary corpus. The Iberian material likewise benefits from having more than one study devoted to it, helping provide a rounded perspective. Isabel Velázquez, who has done more than anyone to explore the world revealed in the slate documents that offer a rare glimpse into the day-to-day economic life of late antique inland Iberia (especially, but not exclusively, the northern Meseta), looks at peninsular Latinity in the Visigothic period. This, she suggests, was a single linguistic code — that is, not Latin on the one hand and proto-Romance on the other — with a single aspirational Latin orthography, but also with a great deal of linguistic evolution visible in the non-literary texts of the era. Vowel sound-shifts were clearly well advanced (though not yet the diphthongisation that would eventually produce bueno from bonus) and it seems likely that many final consonants were no longer voiced though they continued to be learned at school. Gothic was no longer a spoken language by the sixth century, but lexical borrowing and some toponyms suggest that there was a period, no longer directly documented, in which Gothic and Latin were both spoken in some locales and were mutually influential — with the caveat that there is no real way of telling whether a Germanic lexical element goes back to some earlier stage in the coexistence of Latin and the various Germanic languages of the Roman era. Graham Barrett's companion essay, focused on the literary texts from Iberia between 400 and 1000, is a salutary reminder of the obstacles we face in their linguistic analysis. Almost none of them comes to us unmediated — all the manuscripts of the Visigothic law code, for instance, are post-Visigothic, and show clear signs of modification and scribal updating, while early charters mainly survive in high medieval cartularies. Isidore of Seville (‘the one-man renaissance’) is a different sort of challenge, since he presents all his evidence as if Vergil or Ennius were ‘contemporaries and interlocutors’. Nevertheless, Barrett analyses the different linguistic registers of the seventh-century evidence, making particularly interesting use of texts in the middle register, especially the diverse corpus of the Old Hispanic liturgy, which has generally been under-exploited. His observations on the retention of the prolix and frankly difficult Latin of the imperial chanceries in the legislation and epistolography of the period take us back to questions of power — that it was an impediment to comprehension did not lessen the importance of high-flown legal rhetoric as in and of itself evidence of authority. Jonathan Conant's chapter on Africa is something of an outlier, both because it stands on its own and because of the region's later, Islamic history, but among its many virtues is how clearly it demonstrates that ethnicity and language do not map easily onto one another, with the same Saharan immigrants who spread Amazigh languages in the central Maghreb also commissioning inscriptions in Latin.

For their part, the two chapters on the British Isles stand at some remove from anything recognisably Roman, but that very alienness will illustrate for readers of this journal the sorts of questions that must be asked in later periods. Finally, and helpfully, Paul Russell's summary remarks are unexpectedly frank about the methodological and conceptual gaps that the juxtaposition of the volume's papers expose. His own observations on methodology, and on the underlying difference of time-scale between historians’ and linguists’ questions, are another welcome indicator of this collection's value to both groups. It is rare for edited volumes to offer such consistently fine contributions, and even rarer when contributions unrelated to one's own research focus provoke as much interest as they do throughout here.