Roman roads.

      The third reason that Latin spread so widely was the building of roads, a well-known feature of Roman civilization, which in fact extends back to 312 BC, with the first stretch of the Via Appia from Rome south to Capua. This also tended to be a function of the army, since it required a mobile workforce, and was directly useful, above all, to the military. The whole of Italy, and later the whole world north of the Mediterranean, came to be linked by a network of roads that led back to Rome. This was especially true of the western provinces, where Latin became the dominant language. Like the postwar interstate highway system of the USA in our own time, the principal purpose of this government-financed road-building programme was the easy movement of military forces and supplies to wherever they should be needed. But again as in twentieth-century America, the wider community received an immense social benefit, or at least an immense cohesive effect, in making their travel across longer distances more feasible. For all Rome’s new dependencies, the costs of human contact beyond the immediate neighbourhood, for whatever purpose, were lower within Roman domains than they had been before, in their independent past.

      Such enhanced communications gave a differential advantage to Latin: its speakers were able to travel more easily, and farther, than speakers of any other single language. But the road network was also a vast contribution to the prestige of the growing Empire, ostentatiously measured out as it was by mīliāria, ‘milestones’, an extended symbol of pāx Romāna. That in itself acted as a further inducement to solidarity for those who lived within its bounds.

      There is something of a paradox here. Rome’s navy was to defeat every other naval power in the Mediterranean, and largely clear it of piracy as well. Its empire came to occupy the whole Mediterranean coastline and surrounding lands, so that from the second century BC onward the most direct route from Italy to most provinces would have been across the sea. The sheer cost of transporting goods by road must have meant that trade always favoured shipping over wagon haulage: the cost difference was a factor greater than sixty.⁶ Although some of the shorter roads near Rome had indeed had an economic origin (as the Via Salaria, which had supported Rome’s ancient trade in salt with inland communities of Italy to its east, and the Via Latina, which ran southeast toward Campania), it appears that the road system was never primarily an economic infrastructure.

      Yet despite these geographic and economic facts, Roman power, and hence Latin, was never spread by sea power. Rome remained a land-based power, and its roads remained as a highly durable trace of the route marches, itinera, of the Roman army. In a way, the fate of the Etruscans, as well as the limited political success in the long term of the Greek colonies in southern Italy, had demonstrated the lack of durability in what the Greeks would have called thalassokratía and emporía, an unchallenged navy and a trading network. What lasted was a centralized government, a large and mobile strike force, and a readiness to occupy territory permanently.

      Although the long-term outcome is clear, the processes by which the various regions gave up their previous languages in favour of Latin are not well documented. Descendants of Latin are spoken to this day throughout the peninsula, as well as in the offshore islands of Sicily, Corsica, and Sardinia. Latin was already close to universal in Italy by the first century AD. But we can only find traces here and there of the changeover: late inscriptions in an inherited language, transitional effects as inscriptions in one language show the influence of another, the very occasional literary reference to language use.

      What does emerge from this sparse evidence is how long the transition to Latin was delayed. Etruria in fact provides the richest source in Italy of such evidence of language shift, probably because it both had a history of literacy previous to any in Latin and was close to Latium. Epitaphs there have been traced by date, as Etruscan-language inscriptions yielded gradually to Etruscan-Latin bilinguals and then to purely Latin inscriptions.⁷ The spreading wave of Latin-language use passed northward over two centuries, from Caere (25 km from Rome) in the late second century BC, to Volaterrae (200 km farther north) in the late first century AD.⁸

      This means that the progress began as much as eleven generations after Rome’s first military inroad, the destruction of Veii in 396, and a good six generations after the conquest of the last independent Etruscan city, Volsinii, in 264. However, even after the Latin wave had passed, another four or five generations were needed before Roman burial customs were adopted by Etruscan families.⁹ People clearly went on feeling themselves culturally Etruscan long after they had lost their own language, something one can see paralleled in minority language communities to this day.¹⁰

      In this delay, Etruria seems to have been quite typical. Cumae had been the crucial bridgehead for Greek culture into Italy, founded jointly by the Greek cities of Kyme, Khalkis, and Eretria around 750 BC (some 160 km to the south of Rome, at the north end of what would be the Bay of Naples). It had experienced an exciting history of its own before it came under the Roman domination in 343 BC, together with its larger neighbour Capua. Despite being a Greek foundation, its major language was the surrounding vernacular, Oscan. An early and essentially voluntary adherent to Rome’s league, whose citizens had been awarded civitas sine suffragio, Roman citizenship without voting rights, it was well integrated into the Roman system, governed from Rome by annually elected officials (the four praefecti Capuam Cumas, ‘officers for Capua and Cumae’), a post that offered an early rung in Rome’s cursus honorum, the standard career path for a noble Roman. Nevertheless,