“Latin”

The rulers of the Carolingian empire regarded maintaining God's favor as a primary political objective, justifying expansion by conquest and the campaigns of baptism that followed and necessitating a model of governance in which ethical correction played a crucial rhetorical role (Jong; McKitterick, Charlemagne). Because members of the polity belonged to one or another social estate, each of which had its own obligations, a major task of government was to articulate and impose these obligations, by exhortation, imitation, education, or force. This task, which involved a close collaboration between secular and religious elites, meant renewing the polity's commitment to the authoritative examples of the past, a task that in turn required diligent scholarly study of a canon of sacred and secular texts from carefully corrected exemplars.

In an effort to produce the conditions under which this unusually integrated political agenda might be realized, Charlemagne's Epistola de litteris colendis mandates a program of education whose ground was prepared in the Admonitio generalis (General Exhortation), a key legislative work of Carolingian political theology, promulgated in 789 CE (text in Mordek et al.; translation in King 209–20). One section of the Admonitio, drafted by the Northumbrian scholar and royal counselor Alcuin, directs bishops and abbots to establish schools, with the goal of improving the standard of Latinity for the good of the body politic as a whole (Bullough 379–85). The Epistola, also likely drafted by Alcuin, is cutting about the state into which contemporary Latinity had supposedly fallen, declaring in deliberately convoluted syntax that the “fine statements” (“sensus rectos”) that even these same bishops and abbots sent the king were written in dismally “uncultivated terms” (“sermones incultos”). It was “as though sentiments that pious devotion was speaking faithfully within the mind could not be articulated by the unpolished tongue without solecism, because of negligence in study” (“quia quod pia devotio interius fideliter dictabat, hoc exterius propter neglegentiam discendi lingua inerudita exprimere sine reprehensione non valebat”; Charlemagne 233).

Bad Latin was a pressing political concern because it was also a spiritual one, threatening the empire's special relationship with God and thus its survival. In support of its claim that “correct speaking” (“recte loquendo”) cannot be separated from “correct living” (“recte vivendo”), the Epistola quotes Matthew 12.37, “For by your words you will be justified, and by your words you will be condemned” (“Aut ex verbis tuis iustificaberis, aut ex verbis tuis condemnaberis”), implying that, for the professionally literate, learning to speak and write elegant, correct, uniform Latin could make the difference between salvation and damnation in eternity (232).

Latin at the end of the eighth century was both complex and (for some) alarmingly diverse. On a conceptual level, it had long since completed its transition from being only the second most widely spoken language of the Roman Empire, after Greek (Adams, Bilingualism 9–14), to representing itself as one of the “tres linguae sacrae” (“three sacred languages”), along with Greek and Hebrew (Richter). In parts of Europe not dominated by the primarily Greek-speaking Byzantine Empire, Latin was the main language of communication, writing, and Christian worship (Banniard, “Language”). Even in the islands of Britain, which lay outside the direct sphere of Carolingian influence and where textuality was never confined to Latin (Hall), written Old English and Old Irish were in this period used mostly as aids to studying Latin (Blom), which had once been widely spoken in the Roman province of Britannia but was now restricted to an educated elite, of whom Alcuin was one.

For Alcuin and other insular grammarians, Latin was a rich but demanding language of the book, mastery of whose intricacies was a prerequisite for anyone “who aspires to wisdom” (“qui sapientiam desiderat”; Anonymus ad Cuimnanum, line 547; see Irvine 279–82; Law). Similar attitudes were held by other intellectuals who grew up in the Carolingian polity, especially the German-speaking regions east of the Rhine, much of which had been Christianized by missionaries from Britain only recently. In sites such as the monastery at Fulda, founded by the English mission in the 740s and already one of Europe's centers of learning (Raaijmakers), Latin was studied in school using grammars as a matter of course, as in Britain. It was also pronounced phonetically, with careful attention to syllabication, much as it is pronounced today.

However, Latin also had a second identity: as a written form of what would much later come to be called the “vernacular” (Ramminger). Specifically, Latin was the written form of the majority of languages spoken west of the Rhine and south of the Alps, the distant antecedents of modern French, Italian, Spanish, and others. These spoken languages were descendants of the dialectal forms of Vulgar Latin, which had been spoken in Roman times in tandem with formal Latin (Herman; Adams, Regional Diversification). Since the Roman period, the dialects of Vulgar Latin had continued to undergo the ordinary linguistic processes of semantic change and diversification, growing gradually away from written Latin while also becoming steadily less comprehensible to one another (Banniard, Viva voce and “Transition”). By the eighth century, those processes had progressed to a point where spoken communication even between educated speakers from different regions or with different training had become difficult. As the Epistola complains, “sermones incultos”—forms of Vulgar Latin well on their way to crossing the undefinable line between late Latin and early Romance (Vincent)—had also infiltrated the official language of writing, public speech, and worship, creating a growing gap between the Latin that was used and the Latin that was taught.

For intellectuals like Alcuin, for whom Latin was learned, grammatical, and symbolically immutable, any changes to the language were errors, threatening its sacrality, beauty, and political efficacy by rendering it provincial, cutting off region from region and present from past. To prevent disaster, a program of linguistic reeducation was in order. Latin needed to be monolingualized: normalized in the same way preaching and worship across the empire were being normalized, a project central to the Carolingian church. As the emphasis on speaking and word order in the Epistola shows, this meant artificializing the language for Vulgar Latin speakers by standardizing official Latin so that it was written, spelled, and pronounced as it was in Germany and Britain. This in turn meant imposing throughout the polity a single “system of letter-sound correspondences” in pronouncing Latin, “in which every letter had to be given a sound in the liturgy [and other official contexts], and in which in theory every one of those sounds was specified,” as the historical linguist Roger Wright influentially argued (“Late Latin” 351). Only by formally differentiating official Latin from Vulgar Latin could the language again be the instrument of worship and scholarship the empire needed to carry out its sacred political mission. Only by reclassicizing Latin, laying down a path followed six hundred years later by the Italian humanists—often working from manuscripts of ancient texts copied in this period (Witt)—could it retain its status among the “linguae sacrae.”

Whether or not it was quickly implemented in practice, one indication that this monolingualizing of Latin was taken seriously is the countermoves it produced in church legislation, effectively another vehicle of imperial governance. The churchmen who met in council at Frankfurt in 794, soon after the Admonitio generalis and Epistola de litteris colendis had been published, found it necessary to stipulate “that nobody is permitted to believe that God can be addressed only in the three [sacred] languages; God can be worshipped and a person heard in any language, if that person asks justly” (“Ut nullus credat, quod nonnisi in tribus linguis Deus orandus sit; quia in omni lingua Deus adoratur et homo exauditur, si iusta petierit”; Capitulare Francofurtense 171, clause 52). Those who met in council at Tours nineteen years later also found it necessary to instruct priests, preaching in local languages but reading from Latin books, to “be sure to render their homilies lucidly into the local Romance or German dialect, so that what they say can be more readily understood by everyone” (“ut easdem omelias quisque aperte transferre studeat in rusticam Romanam linguam aut Thiotiscam, quo facilius cuncti possint intelligere quae dicuntur”; Concilium Turonense 288, clause 17).

In the excruciatingly demanding Carolingian ecclesiastical model, no baptized person could be without the spiritual ministrations of priests whose own hope of salvation rested on the diligent performance of their duties. As soon as it was mandated, the formalization of Latin required a programmatic rearticulation of the official roles played by local languages in carrying out these duties, whether through preaching, teaching, private prayer, or the veneration of saints (Keefe; Rhijn). As a result, official monolingualism at the apex of the Carolingian church-state bureaucracy necessitated priests who were bi- or multilingual. These priests also had to be adept at text-to-speech translation of professional Latin in didactic genres, whether the majority languages local to their parishes were Germanic or Romance.

Crucially, however, such references to the ecclesiastical use of local languages assumed that these languages were not expected to be part of the textual system. To an extent, this was axiomatic. Because literacy at this period was acquired by studying Latin, there was an intimate alignment between textuality and Latinity. This was true even in Britain, where scholars had developed interlingual study aids in the form of glosses that used standardized sets of English-Latin, Irish-Latin, and Welsh-Latin equivalents (Blom; Lapidge). Despite episcopal concerns about the quality of priestly education, Carolingian churchmen and their successors thus did little to promote written Gallo-Roman, which in practice continued to be understood as akin to a spoken form of Latin until the twelfth century, three hundred years later. The handful of surviving texts that set out to transcribe local Gallo-Roman dialect forms are mostly public performance texts associated with saints’ shrines located in contact zones between Romance and German, where cultic celebration was multilingual and using written Latin would have caused confusion (Ruby). Even the western Frankish secular aristocracy continued to read, and sometimes write, in Latin (Romig).

Nor did Carolingian churchmen or their successors do much as a rule to promote written German, even after the empire's division into linguistically differentiated polities in the late 800s. This point is made explicit in a Latin prologue to an exception to this rule, the rhyming verse Evangelienbuch (Gospel Harmony) by Otfried of Weissenburg. Otfried accuses his compatriots of “laziness” for failing to render “the magnificent splendor of the holy words into our own language” (“divinum verborum splendorem clarissimum proferre propria lingua dicebant pigrescere”; 4), issuing a literary call to arms that was perhaps also a plea for patronage of further works, like the alliterative Christian epic Heliand (Savior), a rare royal commission (Haubrichs).

Although enough early German writings survive from large religious institutions to show that Carolingian Latinity did not actively prevent the textualization of other languages in the empire or its successor polities, even outside Romance-speaking areas it evidently acted as a retardant. Because government and ecclesiastical administration remained monolingually Latinate, there are few German-language examples of writings in the official pastoral genres circulated to priests, where the need for uniformity outweighed other factors.

“Vernacular”

The pragmatic Carolingian Latin of church legislation and priests’ handbooks does not anticipate Italian humanism. But it does anticipate the European organization that grew up in reaction to the Carolingian church-state model during the twelfth and thirteenth centuries, which also made exclusive official use of Latin. This was the Catholic Church, in its new form as a supranational institution headed by the pope, now notionally in complete control of the episcopal hierarchy and the pastoral and sacramental systems it maintained. This intricate institution was served by ever-increasing quantities of scholarly, legal, and administrative texts (Boyle; Clanchy [3rd ed.]) that were themselves only part of a still larger body of texts, many of them historical, encyclopedic, scientific, and literary, that came into circulation in this era of information revolution (Franklin-Brown), on the way toward the media revolution of the fifteenth century, after the invention of printing. The place that Latin continued to occupy in late medieval and early modern Europe as a language of intellectual exchange clearly owes a debt to the Carolingian precedent.

During the centuries before this expansion of Latin texts got underway, however, Carolingian language planning was also playing a pivotal role in shaping a different linguistic development. This was the creation of new bodies of writing in European regional languages, in imitation of Latin but also in complex reaction against its textual hegemony. There were as many versions of this development as there were languages participating in it, all as self-conscious as their Carolingian predecessor, but now expressing their arguments for the textualization of the language in question through literary prologues, not legislation. The urgency of these arguments heralded a politicization of language choice that has remained a permanent feature of European literary culture. Three categories of argument stand out in particular, which I provisionally call the “exceptionalist,” the “imitative,” and the “separatist.”

Apologists for some languages, especially Irish (largely to the west of the Carolingian polities) and Glagolithic (to the east), concentrated on making exceptionalist cases for the status of the language in question, in one case bypassing, in the other expanding and appropriating, the topos of the “tres linguae sacrae.” The seventh- or eighth-century Auraicept na n-Éces (Scholar's Primer), an extraordinarily assertive guide to Irish grammar and script, proposes that the mythical origin of Irish as a learned language created immediately after the fall of Babel allows it to perform functions separate from but equal to those identified with Latin, especially in the discursive realms of secular law and poetry (Ahlqvist; Eco). Far from being inferior to Latin, the two languages are equal, a fact symbolized by the treatise's generic dependence on the tradition of Latin grammars (Hofman), but have separate functions. The ninth- or tenth-century treatise O pismenexƄ (On the Letters), written to defend a newly textualized Slavic language and its liturgy against those who believe that only the “tres linguae sacrae” are accepted by God, also invokes the story of Babel. This text, however, bases its claim for the sacrality of Glagolithic on the language's use of an alphabet derived from Greek, as well as on the holiness of the two brothers who invented this alphabet, the Byzantine philologists Saint Cyril (d. 869) and Saint Methodius (d. 885) (Veder 158–67).

Imitative programs of textualization were less specialized and self-consistent, organized around internal tensions that were as unresolvable as they were productive. Between the ninth and early eleventh centuries, English churchmen, many of them monks working under royal or aristocratic patronage, created the largest body of vernacular writing to survive from any part of early medieval Europe, producing what became essentially a local alternative to Carolingian Latinity. The earliest account of this project, a preface to a translation of Pope Gregory I's Regula pastoralis (Pastoral Care) issued under the name of King Alfred of Wessex around 890, famously announces a program to translate those Latin books “most necessary for all people to know, into the language that we can all understand” (“ða ðe niedbeðearfosta sien eallum monnum to wiotonne, ðæt we ða on ðæt geðiode wenden ðe we ealle gecnawan mægen”; Old English Pastoral Care 8–9) for use in classroom teaching and pastoral care. The preface describes this program apologetically, as a response to the deterioration of English learning because of negligence and generations of Viking raids. But it also justifies the program a different way: as one example of interlingual transmission among many, both ancient and modern. Other examples include the translation of the Hebrew and Greek scriptures into Latin, here seen not as a sacred language but as a link in a chain of learning. Written English is thus both a falling away from Latin and its equal (Godden, “King Alfred's Preface”; Stanton 55–100).

Over the next hundred years, versions of this double justification powered the production of numerous philosophical, historical, biblical, and homiletic writings in Old English, many of them represented as successors to Alfred's translation program (Godden, “Alfredian Project”), in which sociolinguistic ambition and defensiveness remained in strong tension. One result was that English was the first written European vernacular to understand itself as a quasi-official national language, again similar to Carolingian Latin, not least in the standardization of its lexis and orthography (Gretsch). A more durable result was the emergence of a claim, based on the use of English as the language of record for sermons and other pastoralia, that the language was suited only to the “unlearned” (“ungeleordum”), those who did not know Latin. This claim, strikingly at odds with the breadth of the language's actual functions, marks the earliest appearance of a written language in its familiar guise as a “vulgar tongue” (“lingua vulgaris”), whose very accessibility might make it unsuitable for certain texts and topics (Gittos; Watson, Balaam's Ass 261–67).

By comparison with these other languages, French developed a continuous textual tradition late, separating itself from Latin only during the period when the integrated Carolingian church-state model of governance, and the cultural assumptions associated with it, had largely broken down, at the turn of the twelfth century. It was also distinctive in other ways. It emphasized different genres, initially genres of narrative verse; its emergence was notably decentralized, and, remarkably, for the first hundred years of its existence it was written mainly outside the regions where it was the majority language (Careri et al.; Gilbert et al.). Many twelfth-century French texts are commissions by members of provincial aristocracies. Most are also from Norman England, where the majority spoke English, or from another contact zone between Romance and German, such as Picardy (Lusignan, Essai and “Translatio Studii”)—this to the embarrassment of the nineteenth-century philologists who rediscovered and sought to canonize certain of these texts and for whom a primal relationship between language, people, and land was axiomatic (Taylor). These texts thus addressed readerships of mixed linguistic backgrounds and political identities (Gaunt).

Until it began to be identified as a national language during the fourteenth century, written French, much like Persian (Green), was a regional lingua franca, widely used across almost all the former Carolingian polities and beyond, and was lent prestige by an association with Latin that early French writers insisted on by naming the language “romanz” (that is, Roman). Even as this word registers a text's use of dialectal (French) not monolingualized (Latin) forms, the term also asserts the language's appropriation of certain functions of Latin. Latin had long been used in aristocratic and civic as well as ecclesiastical education, uniting elites of different kind and status, in England and elsewhere (Tyler). As Latin became more exclusively identified with the scholastic and ecclesiastical, this role was partly taken over by French, often under the auspices of female patrons, whose role in household education justified the written use of what was called a “mother tongue” (“materna lingua”) for the first time (Wogan-Browne; Grondeux). Even in England, where French was not commonly spoken, “romanz” texts presented themselves as composed in a “common tongue” (“lingua communis”), more accessible than Latin but close enough for equivalence. Early and strenuous efforts were indeed made to expand the new written language's reach by rendering classical and other “subject matters” (“materes”) into French, notably histories and saints’ lives. Even before Chrétien de Troyes placed French in a direct line of inheritance from Greek and Latin in his romance Cligés during the 1170s, a cultural project on some scale was clearly felt to be in progress.

If the Latin of the late-medieval centuries represented one line of descent from Carolingian Latin, French represented another. French soon established itself as Europe's default vernacular, whose versions of Latin texts and genres were usually preferred as sources for works in other vernaculars, especially once Paris became the center of the European book trade in the 1200s. The spread of French and its absorption of many of the functions of Carolingian Latin was an important factor in the narrowing of the domains in which Latin operated. By 1300, Dante Alighieri and others were coming to represent Latin as an artificial language, invented by intellectuals, with no connection to ordinary speech: essential for scholarly purposes but in many contexts no more than a resource for vernacular writers, translators, and their audiences (Botterill). Unlike Latin, French never achieved the hegemonic status of a monolanguage. But it did maintain its geographic reach and cultural prestige, both of them direct inheritances from Carolingian Latin, well into the modern era, reluctantly ceding primacy to English and Spanish only during the last hundred years.

A fuller sketch of the influence of the Carolingian monolingualization of Latin on European language politics over the longue durée would need to incorporate other Romance and Germanic languages (especially Occitan and Norse) and regions (especially the Iberian and Scandinavian Peninsulas). It would also need to push past the thirteenth century to take in the emergence of written Italian, around 1300 (Cornish); the standardization of written French and English over the next century, as these languages came to be represented in nationalist terms (Fisher; Butterfield); and the imitative competitiveness of European languages with Latin and one another that developed as a result (Copeland; Minnis), both before and after the sixteenth-century reformations, when forms of Spanish and Portuguese assumed new identities as imperial languages (Del Valle; Curto) and English was reimagined as the sacred mother tongue of a Christian polity whose church was again headed by a monarch, not a pope (Watson, Balaam's Ass 45–58). It would need to take in even the ostensibly secularized but still numinous accounts of the vernacular as the ground of national identity developed by Romantic philologists in the nineteenth century, a period suffused with medievalism (Rajendran; Watson, Balaam's Ass 49–71), when the transfer of prestige from Latin to Europe's national languages was at last completed, as Western imperialism and the violent acts of linguistic subordination that accompanied it reached an apogee. Benedict Anderson's celebrated 1981 study of European language politics, Imagined Communities, views Latin as an impediment to the development of linguistic nationalism, ignoring both the medieval part of the history of vernacular textuality and its entanglements with Latinity and sacrality. Forty years later, this is one aspect of his justly influential analysis that is clearly unviable.