This was the situation in Ireland from the beginning. Because the island had never formed part of the Empire, the Irish had never known Roman administration, urban life, or organization of schools; rather, they had kept their own traditions and their Celtic language. Still, Latin played a major role in the civilization of this region because of conversions to Christianity at the beginning of the fifth century. In the West Latin was above all the language of the Christian rite and, when Christianity spread beyond the frontiers of the Empire, no one had any intention of replacing it with a native tongue. Latin was also needed for access to the Bible and to the works of the Fathers of the Church. The conversion of the Irish, therefore, led to the need for instruction in Latin on the island. However, this instruction had a limited beginning: it was not intended to train bureaucrats or teachers of rhetoric, but to provide priests and monks with access to Christian literature. To this end an elementary knowledge of the grammar and vocabulary of the new language was necessary, but not a deep investigation of literary texts of the classical era. On the Continent the towns were the centers of culture, where bishops assumed the increasingly greater responsibilities of ancient imperial officials. In Ireland, where there were no towns, ecclesiastical and educational life was concentrated in the great abbeys. The monks studied sacred texts under the direction of the abbot and devoted themselves to the austere asceticism for which the Irish monks were known.

The earliest Latin texts written in Ireland clearly show the result of this peculiar situation. On the one hand, they are full of barbarian and non-Latin features, while on the other, they have a more learned character than contemporary texts written on the Continent. The barbarian side appears, in particular, in the choice of words. Continental authors already possessed a very rich Latin vocabulary in their mother tongue, and, in general, they did not have difficulty choosing the appropriate word. For the Irish, however, all Latin words were equally foreign, and they had to leaf through glossaries to find the desired expression, and, since their reading was limited, the stylistic quality of words escaped them altogether. We find, for instance, in the early hymn Altus prosator, attributed to St. Colomba of Jona (died in 597) rare words such as prosator instead of creator, neologisms such as fatimen and praesagmen derived from fateor and praesagio, Hellenisms such as polyandria in the sense of "sepulchres," Hebraisms such as iduma, "hand." The use of Greek and Hebrew did not mean that they knew these languages. They borrowed these words from glossaries in the same manner as Latin. In the seventh century and later, the Irish often sought unusual words for purposes of rhyme, so that in the hymn Sancte sator, we read: A quo creta cuncta freta, Quae aplustra verrunt flustra, Quando celox currit velox, etc. The unknown author succeeded in collecting an entire series of bizarre forms: creta for creata, aplustra "ships," flustra "calm waters," celox "sailboat." He even appears to have borrowed from the grammar of a colleague the verb geo, derived from e-geo, which was considered to be a compound form: Christo Theo qui est leo Dicam: Deo grates geo (= grates ago).

The exotic quality of this Latin at times stems from the influence exercized by the mother tongue of the Irish. This must be the explanation of the forms staitim for statim, fleatus for fletus, diciabat for dicebat, manachus, Alaxander for monachus, Alexander.

But there also existed a learned and conservative counter-current in the Latin of the island. The missionaries who had brought Christianity had learned their Latin in Roman Britain, perhaps in Gaul as well. They knew how to read or, in other words, they had attended Roman schools, and therefore they brought to Ireland the school pronunciation used in England and Gaul in the fifth century. At that time, many of the changes which we have discussed in the preceeding chapters had not yet taken place. We must also bear in mind the fact that the pronunciation in the schools is always more pedantic and traditional than that of the people. As Ireland was isolated from the Continent certain features of Latin were preserved which the Latin-speaking peoples had abandoned early.

We have, for instance, remarked above that the sound made by the letter c was palatalized before the vowels e and i in the fifth century on the Continent. This change was not yet produced definitively in Gaul and England, when Christianity was introduced in Ireland. The name of the apostle to the Irish, for example, was pronounced Patrikius and not Patritsius; the Irish even today still call him Patrick. The missionaries, therefore, learned to pronounce the letter c as k, even in words such as caelum and civis. This pronunciation became a school tradition in Ireland. For this reason Irish scribes did not write ci instead of ti before a vowel as continental scribes often did. Use of alliteration among the Irish is also very significant. They delighted in tying together as many words as possible in this way in a line of verse, and we have accomplished alliterations in verses where the words begin with the letter c: Clara caeli celsi culmina Cinis, cautus, castus diligentia et Caeli conscendit culmina Caritatis clementia. Even in the twelfth century, the Icelanders visiting Ireland found the pronunciation kelum and kivis, as shown in the first grammatical treatise on the Edda. In this case, for the longest part of the Middle Ages, on the periphery of the world, in a non-Latin land, a linguistic practice stemming from antiquity was faithfully preserved.

Just as important is the handling of endings in the rhymed poetry, where the technique of the Irish differs from the Latin-speakers. In the regions where Latin was used, many phonetic and morphological changes were produced in final syllables. We cannot take up here the complicated history of these changes. It is enough to remark that o and u, e and i were often confused and that the pronunciation of final syllables was weakened especially in northern Gaul (cf., for example, L. vinum > It, Sp vino, Fr vin; L sentit > It sente, Sp siente, Fr sent). When poets began to embellish their verses with monosyllabic assonance, they followed everyday pronunciation and they made short i rhyme with e, and short u with o. So Venantius Fortunatus always makes the iambic dimeters rhyme which he uses in Vexilla regis prodeunt and Agnoscat omne saeculum. We can conclude that for him perfect assonance was formed between the words concinit and carmine, protulit and tempore, praesumeret and debuit, ordinem and ambiit, callido and invidum, redditum and prospero, cernitur and visio, etc. In the same manner, Eugenius of Toledo rhymes suspiriis and conplacet, delectatio and solacium, recogito and transeunt, and Theofridus of Corbie, for instance, rhymes principio with filium, sedibus with versiculos, geritur with gladio. The popular pronunciation is reflected also in the disyllabic rhymes such as fides - crudelis, Christi - estis, adimpleretur -dictum which we find in the poetry of the Merovingian period. There is nothing similar in the Latin poetry of the Irish. They never mix vowels in their rhymes, owed no doubt to the fact that they learned their Latin at school as a foreign language, which they pronounced in their own way, and which they used in establishing themselves on school rules.

The exotic, and at the same time conservative, character of Irish latinity is found to some extent in the ancient Roman province of Britannia. The spiritual and linguistic assimilation of this peripheral province was not yet complete by the beginning of the fifth century, when the Romans summoned their troops to defend the Italian border. The Angles, Jutes and Saxons did not hesitate to invade the region, and they exterminated the Romanized population of the towns and drove back the Celtic population of the countryside further and further to the west. In the land occupied by the Germans, Roman civilization completely disappeared. In the small kingdoms of the Bretons of the west, some remains of the ancient civilization found refuge in Celtic monasteries, where instruction seems to have been organized in the same manner as in Ireland. There at the beginning of the sixth century Gildas lived, the author of a work on the conquest of England by the barbarians. The style of Gildas is inflated and precious, and he is believed to be the same Gildas who wrote the poem Suffragare trinitatis unitas, where the preciosity is carried to an extreme. In this work, the author seeks to protect himself in accumulating formulas of incantation of pagan, rather than Christian, inspiration.

The cultural isolation of Ireland and Celtic Great Britain was interrupted by the pilgrimages of the Celts on the Continent. They preserved their school tradition, their grammatical education and their pronunciation of Latin, but they expanded their horizon and began to study classical literature, traces of which are already discernible in the writings of Colomban (d. 615).

Before this development, the Celtic and Roman civilizations met and clashed with one another in Germanic England. In the hands of the barbarians the region was Christianized early and was reclaimed for civilization by two groups, monks coming from Ireland and Roman missionaries. At the beginning of the seventh century, the Irish founded several important monasteries, for example, Lindisfarne and Whitby in the north, and Malmesbury in western England. In these abbeys an Irish type of education was given to the Anglo-Saxons, who adopted the Irish pronunciation of Latin, among other things, and preserved it for a long time. It is likely that the Venerable Bede and Alcuin pronounced ce and ci as ke and ki. We can draw this conclusion from their use of alliteration. So, Bede regularly uses two alliterations in each line of his hymn which begins with the strophe:

Adesto, Christe, cordibus,
Celsa redemptis caritas,
Infunde nostris fervidos
Fletus, rogamus, vocibus.

We have an alliteration between Christe and cordibus in the first line, between celsa and caritas in the second, between in-funde and fervidos in the third, and between fletus and vocibus, pronounced focibus, in the fourth (see below). In his poem Nunc bipedali, Alcuin tied Adonic verses two by two with an alliteration of this type:

Esto paratus ecce precamur
Obvius ire omnipotenti
Pectore gaudens Pax tibi semper. . .

It is, therefore, likely that he pronounced kerte in the same manner as kurva in the two lines: Curva senectus certus propinquat. The Anglo-Saxons kept this pronounciation into the tenth century. When Abbot Fleury lived in the convent of Ramsay in England between 986 and 988, he composed a little book entitled Quaestiones grammaticales, in which he criticizes the pronunciation ke and ki. He writes: Quod quam frivolum constet, omnibus vera sapientibus liquet. For him the pronunciation tsivis, which he learned in Gaul in his youth, was nice and proper, while kivis, which he learned in England, was barbaric. It did not occur to him that, actually, the barbarians had preserved an ancient usage which the Latin-speaking peoples had abandoned.

Nevertheless, Anglo-Saxon civilization would not have achieved its splendid rise in the time of Bede and Alcuin, without the influence of Rome. In 597, Gregory the Great sent the monk Augustine to Canterbury to preach the gospel to the barbarians, a mission which was to bear extraordinary fruit. The penetration of Roman influence to the north and west brought on a clash of Roman and Irish interests. The conflict lasted some decades. But in 669, Pope Vitalian decided to send Archbishop Theodore, accompanied by the monk Hadrian, to Canterbury, to organize the church in England. Theodore was originally of Tarsus and had received his education in the Greek east. Hadrian, who also knew Greek, came from Africa where the ancient Roman school-system was still active. Both knew profane literature as intimately as Christian, Greek as well as Latin, if one can accept the word of Bede. At the episcopal school and the monastic school of Canterbury, Theodore and Hadrian brought together a cotery of students who learned, among other things, metrics, astronomy, computus and who, according to Bede, pursued the study of Greek and Latin to the point where they spoke these languages as well as their mother-tongues. We can confirm that his judgement is correct as far as Latin is concerned. As for Greek, the knowledge of the language among the English was never profound and disappeared with the students of Theodore and Hadrian.

The first group of Anglo-Saxons taught at Canterbury still had close ties to the Irish tradition. This was the case with the first Anglo-Saxon author, Aldhelm. Before being the student of Hadrian at Canterbury, he had been taught by the Irish Maeldubh, who directed the abbey of Malmesbury in the middle of the seventh century. The Latin of Aldhelm presents a two-fold aspect. Rare words, borrowed from the glossaries, and the inflated style recall the "hisperic" Latin, which we have discussed. On the other hand, the linguistic certainty and the wide reading of Aldhelm come chiefly from his studies at Canterbury.

The following generation brought Latin culture in England to its height, a result of the new contact with classical authors. In the kingdom of Northumbria, Benedict Biscop had founded the great abbeys of Wearmouth and Yarrow between 674 and 685; to these he gave an important library of manuscripts, brought from Rome. In the midst of these books the Venerable Bede grew up, first at Wearmouth, then at Yarrow. Bede is perhaps the greatest scholar of the central Middle Ages. He handled the Latin language with remarkable ease drawing inspiration from ancient authors; his style is clear, simple and easy to understand.

The same humanistic activity animates the episcopal school of York, directed between 686 and 721 by John of Beverley, an early student of Theodore of Tarsus. Alcuin, born about 730, lived there until 781, when he met Charlemagne while on a trip to Italy.