Vulgar Latin as an emergent concept in the Italian Renaissance (1435–1601): its ancient and medieval prehistory and its emergence and development in Renaissance linguistic thought

4.3.1 Biondo 1435

In his treatise De verbis Romanae locutionis (1435), the historian Flavio Biondo (1392–1463) explores the identity of the language of Roman oratory (orationes concionesque) and of general language (universa locutio, Biondo 1435: § 6). The Renaissance humanists’ recovery of Cicero’s previously unknown speeches brought to the fore the question of the language used by Roman orators in their public speeches: whether it was the contemporary Italian vernacular (aetate nostra vulgata idioma) ^[21] or the grammatical language (grammaticae artis usus), ^[22] which is called Latin (quod Latinum appellamus, Biondo 1435: § 8). The recovery of Cicero’s and Quintilian’s rhetorical treatises (De Oratore, Orator, and Brutus and Inst. orat.) provided a key to answering this question. Biondo opposed the theory of bilingualism, arguing that there was in antiquity only one single Latin, that is, one single linguistic unity of Latin, regardless of social class, gender, degree of literacy and formal education (Biondo 1435: § 6, § 37, § 64, § 86, § 89). However, this theory of monolingualism is not a theory of linguistic uniformity. Biondo recognized internal variability of ancient Latin both on the individual level, according to shifting linguistic ability (diversa dicendi facultas, Biondo 1435: § 64), and in functional levels, according to a tripartite socio-cultural division.

Quoting Cicero, Orat. 195, Biondo (1435: § 39) uses the criterion of rhythm to classify ancient Latin into three forms of diction (tres latinae dictionis formae): a poetical (forma poetica), an oratorical (forma oratoria), and a popular form (forma vulgaris). ^[23] The two former represent learned Latin, and the latter unlearned Latin. He applies a historical perspective to the emergence of these forms from the common stage of Latin (par forma, communis dictio). Through the practice of modulation and public speaking (Biondo 1435: § 45), the first and second forms achieved a prestigious status (primariae). Owing to the neglect of cultivation among common people (inter vulgares), the third form was to remain artless, rude and disreputable (incompta, squalida, ingloria). Biondo defines forma vulgaris as lacking rhythm (sine numero, Biondo 1435: § 39), just as Cicero had defined sermo vulgi (see Section 2.1 above). He adds that it lacked artfulness and rules (nulla arte, nullis habebatur regulis, Biondo 1435: § 42) and was used by common people (in vulgi possessione, Biondo 1435: § 42), for which reason it was known as vulgaris. It was not devoid of the quality of latinitas (Biondo 1435: § 43), but rather consisted of the same words as the two other major forms of diction (Biondo 1435: § 42).

Biondo discussed the lexicon (Biondo 1435: §§ 47–71), prosody (Biondo 1435: §§ 82–90) and grammar (Biondo 1435: §§ 91–100) of forma vulgaris in relation to his contemporary Italian (nostra vulgare). As parameters of his analysis Biondo used the concepts of latinitas ‘(correct) Latin’ and vulgaritas ‘vernacularity’ (Biondo 1435: § 32, § 86, § 107) with their variants: latinus and vulgaris (Biondo 1435: § 47, § 78) latine and vulgariter (Biondo 1435: § 21). In ancient grammatical theory, latinitas in its normative sense means correct Latin on three different levels: phono-lexicon, prosody (that is, vowel quality, intonation, and stress), and morphosyntax. Biondo’s application preserved these levels, but modified the concept of latinitas. Instead of using it in its ancient sense, to designate correct Latin, Biondo used it to designate Latin as a linguistic type in contrast to other languages, in this case Italian. This conceptual modification or transformation of latinitas was necessitated and prompted by Biondo’s object of comparison, namely contemporary Italian. This change from a normative to a crosslinguistic sense of latinitas made it a useful tool for dealing with the question at hand. ^[24]

In general, Biondo argued that the speakers of the forma vulgaris attained lexical latinitas, could recognize quantities of syllables in versified speech, ^[25] and came close to grammatical latinitas. Concerning the Latin or Italian lexical nature of forma vulgaris, Biondo argues on the basis of the practice of Roman orators. He argued that the orators used Latin words (Biondo 1435: § 47, § 64, § 68 Latina verba) that are dissimilar to words of the contemporary Italian vernacular (verbis nostris … nequaquam adsimilia, Biondo 1435: § 47). The lexical register varied from high to plain style, depending on audience (Biondo 1435: § 41, § 47, §§ 69–70). Regardless of the stylistic level of words, the words of public speeches were Latin (Biondo 1435: § 68). Concerning the grammatical ability among users of forma vulgaris, Biondo argued that uneducated people did not know the rules of grammar by nature or by good usage (nec naturae aut bonae consuetidine munere). Though they deviated from grammatical latinitas in morphology and syntax (Biondo 1435: §§ 91–93), they were not too far from it. To confirm his position, Biondo claimed that the quality of grammaticality is intrinsic to vernaculars (Biondo 1435: § 100 idiomatis natura insitum), even among users of the most corrupt vernaculars (corruptissima etiam vulgaritate loquentibus, Biondo 1435: § 100, cf. Coseriu and Meisterfeld 2003: 155).

Biondo (1435: §§ 72–73) explicitly refrained from discussing his topics from an exclusively rhetorical perspective, that is, from within the domain of ornatus in classical rhetorical theory. He states that he confines his perspective to the phono-lexical form of the Roman language, as opposed to its interior syntactic and stylistic structure (Biondo 1435: § 72 verba vs characteres, locutio vs compositio, cortex vs medulla). Biondo’s reason is that the lexical base of Latin in classical antiquity was almost always the same, or at least semantically comprehensible to everyone, regardless of the degree of its rhetorical elaboration in the three stylistic levels (the grand, middle, and plain, Biondo 1435: § 73).

To conclude, Biondo proposed the concept of Vulgar Latin as a historical socio-situational variety of ancient Latin designating artless, unlearned usage, as opposed to elaborate, learned usage. In reviving and modifying the concept of latinitas, he developed a lexico-grammatical perspective, in accordance with the specific levels or contents of latinitas in ancient linguistic theory (viz. phono-lexicon, prosody, morphosyntax). Tavoni (1986: 27) argues that Biondo has a rhetorical perspective. Biondo himself claims to have a non-rhetorical perspective.

4.3.2 Poggio Bracciolini 1450

In his Disceptatio convivalis, book 3, Poggio raises the question of whether the ancient Romans (prisci Romani) spoke Latin as their natural vernacular and whether the language of educated people was distinct from the popular language. Poggio advances different kinds of arguments. First, he proposes an ethno-linguistic proof of the Latin identity of the language spoken by the ancient tribe Latini:

The language that was in common use among the Latins ought to be Latin. […] Consequently, from the Latins the language is called Latin, which they, not others, used by then. (Poggio 1450: § 14 and § 17)

Secondly, he argues from a collection of evidence from ancient authors (mainly Cicero, De oratore, Orator, Brutus, Quintilian, Varro, Gellius) that all ancient Romans, even uneducated people (vulgus imperitum), spoke Latin publicly and privately, and thus used Latin as their vernacular (Poggio 1450: §§ 29–126). Poggio calls attention to passages on the quality of latine loqui in ancient linguistic theory and on the use of Latin words in the context of spoken language. As regards references to latine loqui, he seems, like Biondo, to construe them in a crosslinguistic contrastive sense, as opposed to Italian, rather than in their original normative sense.

From Quintilian, Inst. orat. 1, 6, 27, on the alternative inflection of senatus (-i and -us) and the distinction ‘to speak Latin’ (latine loqui) versus ‘to speak grammatically’ (grammatice loqui), Poggio concludes:

From this it is clear that everyone spoke Latin, but not everyone spoke grammatically […]. Consequently, everyone spoke Latin, but the learned more correctly, emendatius. […] For the learned judged by systematicity (ratio), what the others attained by usage (usus). (Poggio 1450: §§ 38–39)

Poggio argues that ordinary speech and eloquent speech did not differ in lexicon, but only in rhetorical ornamentation. Quoting Quintilian, Inst. orat. 12, 10, 43, Poggio concludes:

From this it is clear that the language of common and eloquent people was the same as regards the semantic precision (sermonis proprietas), but that it differed in rhetorical style (elocutio) and embellishment (ornatus). (Poggio 1450: § 41)

When commenting on lexical differences of style in contemporary Italian, Poggio returns to this diaphasic view of linguistic variation in ancient Latin:

The same may also have been the case in the Latin language, namely that although the words were common to all (communia omnibus verba), the eloquence and elegance of the words were not so. (Poggio 1450: § 74)

To conclude, Poggio revived the ancient concept of sermo vulgaris as a diaphasic variety on the basis of his interpretation of Quintilian.

4.4.1 Francesco Filelfo (1451–1473)

In his letter to Francesco Sforza (1451), Filelfo (1398–1481) continued the debate between Biondo and Leonardo Bruni on the linguistic situation of ancient Rome (Filelfo 1451: §§ 1–2). Filelfo suggests that the linguistic situation in the late Republic – from Plautus and Terence to Cicero and Caesar (Filelfo 1451: § 3) – was monolingual, but at the same time was marked by a difference between ordinary or common Latin (oratio forensis, popularis, Filelfo 1451: § 4; communis et quotidianus loquendi usus, Filelfo 1451: § 8) and literary Latin (sermo literalis), the language of poets and philosophers (Filelfo 1451: §§ 5–7; cf. Tavoni 1984: 174–175). To demonstrate this difference, Filelfo makes a typological comparison with other speech communities. He claims that the linguistic situation of ancient Rome was similar to the state of diglossia among Jews, Greeks, Egyptians, Arabs, Syrians, and Persians in his own time. He underlines that the divergence between ordinary language and literary language does not make them different languages; the differences instead make up a single linguistic system (Filelfo 1451: § 9):

I know that Jews and Greeks, and I have learned that Egyptians, Arabs, Syrians, Persians, and in the same way all other nations, who have a particular grammar, when employing one literary language, sermo litteralis, and another vernacular, sermo vulgaris, the literary language is more refined (limatior) and possibly more semantically precise (proprius) than the mother-tongue, but is still not another language (non alius) nor completely different (omninoque diversus) in the inflectional variation of words (dictionum variatio), in declension (declinatio), pronunciation (enunciatio) and desinence (desinentia, i. e. use of inflectional suffixes, termination). (Filelfo 1451: § 9) ^[26]

Filelfo’s statement may be interpreted as a claim of a weak diglossia in Ferguson’s sense (see Section 1.2 above, footnote 5). Such an interpretation is confirmed by the historically well-known diglossia of the Greek, Arabic, and Persian language communities in the Middle Ages and the early modern period.

In the letters to Maria Blanca (1463) and Lorenzo de’ Medici (1473), Filelfo contrasts Latin in general (latinitas, Latinus sermo) with literary Latin (literatura, ^[27]literalis sermo) in the time of the late Roman Republic (from Plautus and Terence to Cicero and Caesar):

The Latin language at that time was very close (admodum proxima) and similar (persimilis) to the literary language (litteratae), but now it is so different (dissimilis) that it appears to have become completely barbaric. (Filelfo 1463: § 3 [De Keyser 2015 (2): 908, lines 33–34])

Are the literary language and the Latin language (sermo literalis et latinus) one and the same or entirely diverse? They are neither the same nor different in every respect (nec ex omni parte alius). (Filelfo 1473: § 12)

Just as it was one thing to speak Attic and another to speak grammatically among the Athenians, the same difference also existed among the Romans, so that the principles of Latin (latinitatis ratio) and of literature (literaturae) were different, but this difference was however very small (sed ea tamen admodum parva). (Filelfo 1473: § 59)

Filelfo exemplifies this with heteroclitic words. Twofold nominal inflections according to the 4th or 2nd declension (e.g. senatus with -i or -us) and the 3rd or 1st declension (barbaries, -ei vs barbaria, -iae) are referred to as constituting instances of the difference between Latin and grammatical/literary inflection (Filelfo 1473: §§ 59–60).

Most importantly, Filelfo does not coin the term Vulgar Latin, but only describes general Latin with a number of qualifications: “vulgaris et omnibus notus” (Filelfo 1451: § 18), “communis et quotidianus loquendi usus” (Filelfo 1451: § 8), “doctis indoctisque communis” (Filelfo 1473: § 12), “materna ipsa vernaculaque lingua” compared to Greek koiné (Filelfo 1451: § 12), “quotidiana vulgarisque” (Filelfo 1473: § 17), “vulgaris et omnibus communis” (Filelfo 1473: § 41), and “vulgaris et omnibus cognitus” (Filelfo 1473: § 63).

He recognizes a normalizing effect of literary Latin on general Latin, when the latter becomes corrupted:

I mean, the Latin language was popular (Latinus sermo vulgaris erat) and known by all et omnibus cognitus), but the literary language (litteralis) was not at all so, and was only known by the experienced and learned, however it was of such a nature that it corrected and nourished corrupt Latin. (Filelfo 1473: § 63)

4.4.2 Lorenzo Valla 1452, 1453

In De linguae Latinae elegantia, Lorenzo Valla (1407–1457) attempted to restore the ancient Roman language on the basis of the usage of prose writers, namely the greatest philosophers (summi philosophi), orators (summi oratores) and jurists (summi iurisconsulti). This restoration aimed to bring about a revival of Roman culture. He mainly comments on morphological, lexico-semantic, and phraseological features.

In two fictional dialogues between himself and Poggio, Antidotum in Pogium: liber II (Valla 1452) and Dialogus: libellus secundus (Valla 1453), Valla tries to refute Poggio’s position in Historia convivalis. These dialogues take the form of a metadiscursive criticism of Poggio’s position and a re-interpretation of Bruni’s position. Valla and Poggio held different opinions on whether ancient Latin was an artificial language of the learned or a living language of the common people. Valla accepts the medieval idea of Latin as grammatical and vernaculars as ungrammatical (Valla 1452: § 195, 1453: § 50, §§ 80–81). To refute Poggio’s position, Valla resorts to the strategy of employing the term Latin (latine loqui) in two different senses: either (i) in a stylistic sense, referring to observation of the usage of learned people (consuetudo peritorum) as an alleged rhetorical virtue, or (ii) in a crosslinguistic contrastive and typological sense, according to which Latin is the language both of the ancient and modern Romans, so Italian can be called modern Latin. Valla does not consider Latin in its normative sense (correct Latin). Valla takes latine loqui in Quintilian (Inst. orat. 1, 6, 27) as an instance of the stylistic sense of Latin, as a quality of the orators and the learned in their conformity to rhetoric (Valla 1453: § 38), in contrast to Poggio, who interpreted it as referring to the Latin vernacular in a crosslinguistic, contrastive sense, as compared with Italian. After that, Valla in his dialogue switches to a typologically contrastive use of the term Latin:

The interlocutor Poggio says: “Have I not said that all Romans spoke Latin (latine), but the learned more correctly (emendatius, § 43)?”

The interlocutor Valla answers: “I do not understand why you say ‘more correctly’ (emendatius). I admit that some people may speak more ornately, in a more lofty way, and more eloquently than others, but not ‘more Latin’ (non latinius), if all the Romans had the Latin language in common (communis lingua latina, § 44). […] If the ancient Romans (prisci Romani) did not speak in more Latin way (latinius), mutually, it amounts to nothing, what you are saying, that the learned spoke more correctly (emendatius, § 45)”. (Valla 1453: §§ 43–45) ^[28]

Commenting on Quintilian (Inst. orat. 12, 10, 43), Valla writes: “Behold what Leonardo stated, that there was one popular (vulgaris) and another learned language (eruditus sermo), something you completely deny. If this difference was great, Leonardo was right; if it was small and almost nothing, you win” (Valla 1453: § 21). Thus, Valla redefined the difference as a question of difference in degree rather than in nature, but he does not express his opinion of the degree of this divergence, i.e. whether it was great or weak, despite declaring such a purpose (Hoc ipsum quantum distantiae erat querimus, i.e. ‘Let us explore this, how far the distance was’, Valla 1453: § 77). By analogy with contemporary Italian, however, he indicates that the high variety differed greatly from the low variety in phonology and in grammatical features, such as the use of definite article, declension of nouns, (present) participle, gerund, supines and inflection in general (derivandi usus). The lack of such features in contemporary Italian serves as the basis of the implicit assumption that they did not occur in the ancient low variety (Valla 1453: § 83, §§ 85–86). Valla avoids expressing his own opinion, since his stated purpose is only to oppose Poggio’s opinion (Valla 1453: § 87).

Valla asserts that there is a historical continuity between popular Latin and the Italian language of his own time (Valla 1453: § 25, § 73). Poggio had made a terminological distinction between Romana lingua as modern Italian and Latina lingua as ancient Latin. Valla retains the ancient binomial whereby lingua Latina is equal to lingua Romana both for ancient Latin and for contemporary Italian (Valla 1452: 195). Thus, Valla combines his notion of an ancient diglossia with his diachronic perspective on the continuation of Latin until his own time.

Why did Valla not state or develop a clearer concept of and terminology for Vulgar Latin? The answer seems to be that his humanistic Latin program of language education and the prevailing situation of language policy dissuaded him from assuming too great a distinction between the popular Latin language and the literary Latin of antiquity for the following reasons:

1. His program for a revival of Roman culture was based on a single Latin language system (“As many nations have one single code of laws, the Roman language is single”, Valla writes in the preface to De linguae Latinae elegantia, book 1).

2. There was a socio-political danger with acknowledging a Latin vernacular tradition independent of classical Latin since antiquity, namely that Rome would run the risk of losing the competition for linguistic superiority with Florence (Mazzocco 1993: 77–79; Nauta 2009: 278).

4.5.1 Guarino of Verona 1449

In a letter to Prince Leonelle d'Este, De lingue Latine differentiis, Guarino of Verona (1374–1460) considered the nature (vis) of ancient Latin, its diachronic change (differentia) and its decline (ruina, Guarino 1449: § 6) in order to answer the question “what was the language of the Roman ancestors of the Italians: contemporary Italian or grammatical Latin”? Guarino distinguished between two types of Latin, based on their method of acquisition: (i) Latin in a non-normative sense, and (ii) Latin in a normative sense (Guarino 1449: § 8). In the former sense, Latin was learned by usage without method and rules (sine ratione, sine regulis) and was used by rural and urban people (urbani et rustici). It conformed to its own grammatical regularity, as an inherent and natural quality, among users devoid of formal grammatical education (Guarino 1449: § 46), in analogy with contemporary vernacular Byzantine Greek (Guarino 1449: § 45). Thus, uneducated persons were able to speak and understand Latin well (bene, Guarino 1449: § 34). In the second sense, Latin was acquired by formal training (studio et arte), and this refers to correct Latin, i.e. Latin without phonetic faults (barbarisms) and morpho-syntactical errors (solecisms, Guarino 1449: § 9). Guarino despised the contemporary Italian vernacular (materna, plebeia, vulgarica), considering it unworthy of the name of Latin. This testifies to the survival of the medieval use of the term Latin to refer to the vernaculars, a use which the Latin humanists tried to eradicate in their endeavor to revive classical Latin.

In his survey of the internal historical changes of Latin in antiquity (Guarino 1449: §§ 12–21), Guarino classifies ancient Latin into four successive stages (viz. lingua prima, lingua secunda, lingua tertia, lingua quarta), drawing on Isidore’s periodization of Latin into prisca (proto-Latin, 1200–850 ad), Latina (Latin in the Kingdom of King Latinus, from ca. 850 bc), Romana (Republican Latin, from 509 bc), mixta (from the time of Roman Empire at its high point of expansion). As regards the pre-republican stage, Guarino says that people learned Latin through usage (consuetudine) rather than through formal training (ratione et artificio, Guarino 1449: § 17). Otherwise, Guarino only aims to prove the theory of monolingualism, which implies that in the pre-republican and republican periods the entire population spoke and heard Latin (Guarino 1449: § 22, § 52) in contrast to its restricted use among the learned in his own time (Guarino 1449: § 33) after the language deteriorated as a result of interference from the language of Goths and Langobards (Guarino 1449: § 21 and § 52).

In general, Guarino understands the changes in Latin during the first periods as a development towards cultivation and refinement. The culmination in the third period (Guarino 1449: §§ 18–19) was followed by a general decline caused by the influence of foreign nations speaking Latin incorrectly (inquinate loquentium, Guarino 1449: § 21). As Mazzocco (1993: 55–56) notes, Guarino agrees with Biondo in arguing in favor of a monolingualism of Roman antiquity, but displays a greater sensitivity to the chronological development of ancient Latin, though without precisely relating it to the two types of Latin – unlearned and learned – which he had identified.

5.4.1 Gabriel Barrius, 1554

Gabriel Barrius (1506–1577), a Latin humanist and historian from Francica in Calabria, is a less known Renaissance scholar, who endeavored to defend Latin in the Italian debate on language cultivation (questione della lingua) in his three-volume Pro lingua Latina (1571 [1554]). He also wrote a historical-geographical study of Calabria, De antiquitate et situ Calabriae (1571). In the former treatise, Barrius defends the position of ancient and humanist Latin against the views of Bembo, who had relegated Latin to the status of a foreign and non-domestic language.

To defend Latin’s position as a natural and domestic cultural legacy of Italy, Barrius within the framework of corruptio, developed a concept of Vulgar Latin as a low variety (latinus vulgaris sermo, quotidianus vulgaris sermo) that deteriorated into Italian because of contacts with the Goths, while the high variety (Latina regularis lingua, grammaticus latinus sermo) remained unspoiled and undamaged (Barrius 1571 [1554]: 135, 170–172). To defend Latin as a national language, Barrius raises and answers a rhetorical question:

If the Latin vernacular from Latium is ancestral to the Latin and Roman nations as the native, indigenous, domestic, everyday, common, public, popular language (vulgaris sermo), why is grammatical, literary, regular, artistically produced Latin not ancestral to itself? […] Accordingly, literary Latin (Latinus literalis sermo) is definitely ancestral (patrius) to the Latin nations, native, our own and peculiar to us (nativus, proprius et particularis). (Barrius 1571 [1554]: 99).

All metalinguistic attestations of popular language (sermo vulgaris), popular words (verba vulgaria), words of common people and the rural population (vulgi et rusticorum vocabula), and everyday language (quotidianus sermo) refers, according to Barrius, to one and the same linguistic entity, Vulgar Latin (latinus sermo vulgaris, Barrius 1571 [1554]: 226). Barrius points out that the concepts of ordinary language (vulgaris loquendi modus) and elevated language (singularis loquendi modus) are relational. He qualifies ordinary language with a number of adjectives, viz. everyday, popular, public, and common to everyone (quotidianum, populare, publicum, et commune omnibus, Barrius 1571 [1554]: 226). He regards elevated language as artful, rare, prominent, and not common by nature to everyone (artificiale, rarum, excellens, non omnibus natura commune, Barrius 1571 [1554]: 226). Barrius repeatedly qualifies vulgaris Latinus sermo with a long set of adjectives, viz. native, vernacular, everyday, familiar and popular (nativus, vernaculus, quotidianus, familiaris, popularis, Barrius 1571 [1554]: 235, cf. 99, 168). He uses all these qualifications in order to represent spoken Latin as a national Italian language and thus to oppose the claims of Bembo (Section 5.3).

Furthermore, Barrius transformed the medieval theory of bilingualism into a theory of diglossia in Ferguson’s sense (Section 1.2, footnote 5), apparently a high-level one. In the preface, he certainly describes his work as a search for the existence of a bilingual co-existence of Latin, Latinus sermo, and the vernacular, vulgaris sermo, since antiquity, but in the course of his treatise he applies the term Latin to both of these entities, which means that he presupposed the existence of two varieties of one and the same language, namely Latin, rather than two different languages, as Laureys (2010: 30) observes. This clearly means that Barrius presumed that a state of diglossia (in Ferguson’s sense) existed in ancient Rome. The medieval notion of Latin as grammatical and the vernacular as ungrammatical pervaded his linguistic thought (Barrius 1571 [1554]: 11–12; Laureys 2010: 32). His didactic views on written and grammatical Latin as having been a second language ever since antiquity guided his theorizing, because mastering Latin has always required formal study.

Barrius clearly underlined the qualitative difference between the high and low varieties. ^[34] Like Valla (1453: § 85; cf. Section 4.4.2), Barrius (1571 [1554]: 169) thought that Vulgar Latin, as a linguistic type, lacked noun declension, verb conjugation, participles, gerunds, supines, degrees of comparison, and neuter gender. Concerning the lexical stock, Barrius regarded a great number of words as originally common to both Vulgar Latin and standard Latin (latinitas), including amo, curro, fabrico, subito, quando, supra, intra, intro, infra, poeta, via, casa, porta, and iustitia. On the other hand, standard Latin possesses many words not known to Vulgar Latin, and vice versa. Authors of the Latin language took rude and formless words from common popular usage, formed and fashioned them, and endowed them with Latin citizenship (Barrius 1571 [1554]: 176, 190). They also formed new words from extant roots, such as perago from ago, perlego from lego, and nauta from navis. Barrius perceived the difference between Vulgar Latin (Latina vulgaris) and grammatical Latin (Latina regularis lingua) as a relation between natura and ars, comparing the former to a rough stone and the latter to a bas-relief (Barrius 1571 [1554]: 190). In several passages, Barrius comments on the historical relationship of Vulgar Latin to regular or grammatical Latin. He traces the high variety back to the early Kingdom (Barrius 1571 [1554]: 132, 164).

Barrius also justifies his theory of the diglossic situation of ancient Rome with reference to the similar situations in the Greek and Jewish speech communities. They have two kinds of language (duo loquendi modi), an ordinary one, common to everyone, and an elevated one, which he equates and identifies with regular language. ^[35]

5.4.2 Castelvetro 1572

The Italian philologist Lodovico Castelvetro (1505–1571) explained the historical origin of Italian as an internal development from Vulgar Latin. In doing so, he formally adhered to the corruption-generation hypothesis, but in fact he abandoned it (Faithfull 1953: 286). In his notes (Giunte) on Bembo’s Prose della volgar lingua (1572), Castelvetro expounds his views on the nature and variation of ancient Latin (Castelvetro 1714 [1572]: I, 18–31) and on the transformation of ancient Vulgar Latin into Italian (Castelvetro 1714 [1572]: I, 31–37).

In his criticism of Bembo, Castelvetro developed the diglossic concept of Vulgar Latin. In postulating a diglossia (in Ferguson’s sense) between noble, cultivated, and literary Latin as a high variety and popular Latin as a low variety, Castelvetro takes an intermediate position between the theories of bilingualism and monolingualism. In doing so, he interprets Leonardo Bruni’s position in an unbiased way, well aware of the different conclusions drawn by Filelfo and Poggio on the one hand, and Valla on the other (Castelvetro 1714 [1572]: I, 26).

In relation to the medieval bilingualism theory, Castelvetro thinks that the grammar of Vulgar Latin differed from that of the Italian of his own time (“la lingua de’ nostri tempi, chiamata volgare”, Castelvetro 1714 [1572]: I, 28), but that its lexical stock was similar. Castelvetro posits two social varieties: the language of orators, historians and poets and all writers of literary ambition, as opposed to the language of the people (“la lingua antica del volgo”, Castelvetro 1714 [1572]: I, 29), spoken by people of the low classes and farm-workers (Castelvetro 1714 [1572]: I, 28–29). He also addresses the question of the mutual relationship between these two varieties. In his view, the literary variety was not so far from the popular one. They had grammatical features and a large number of words and phrases in common. Castelvetro agrees with Poggio and Filelfo that written Latin was generally understood by the people (“la lingua latina scritta fosse intesa dal popolo universalemente”, Castelvetro 1714 [1572]: I, 29), but underlines the historical continuity of ancient popular Latin, in most of its lexicon, with the contemporary language (“presente linguaggio”, Castelvetro 1714 [1572]: I, 29) and its terminological continuity as volgare (Castelvetro 1714 [1572]: I, 29). To quote Marazzini, “Castelvetro arrivava ad una formulazione abbastanza precisa del concetto di ‘latino volgare’” (Marazzini 1999: 72). ^[36] To express this concept, Castelvetro employs a definite term “la lingua latina volgare” (Castelvetro 1714 [1572]: I, 43), previously attested in Barrius in its Latin counterpart, but his terminology is not uniform. ^[37]

Addressing the question of the emergence of Italian, Castelvetro (1714 [1572]: I, 31) says that all that remains to investigate is when and how Vulgar Latin diffused and changed. In doing so, he divides the process into three periods: (i) before the Sack of Rome 410 ad, (ii) the rule of Goths in Italy, (iii) the rule of Langobards in Italy. In the first period, Vulgar Latin spread into the usage of higher classes. This social diffusion was caused by emperors and courtiers of foreign, non-Italian origin, who did not communicate in pure and literary Latin, but in Vulgar Latin (Castelvetro 1714 [1572]: I, 32). At the time of the Emperors Honorius and Arcadius, Vulgar Latin had spread to all the population in Italy (Castelvetro 1714 [1572]: I, 33). In this period, Vulgar Latin was uncorrupted and preserved its original grammatical features (“suoi primièri accidenti”, Castelvetro 1714 [1572]: I, 33). In the second period, contacts with the Goths had only a limited effect on the Latin vernacular. People used an uncorrupted Vulgar Latin (“la latina lingua intera”) until the Langobards arrived in Italy (Castelvetro 1714 [1572]: I, 34).

In the third period, Vulgar Latin was corrupted in the course of the first hundred years after the arrival of the Langobards (Castelvetro 1714 [1572]: I, 35). It underwent grammatical changes (“mutamenti accidentali”, Castelvetro 1714 [1572]: I, 35) and began to take the form that it still has. Germanic grammar had a functional influence on Vulgar Latin, so that extant elements were modified to fit the usage of Germanic nations (e.g. the use of articles, the conjugation system (Castelvetro 1714 [1572]: II, 55–56 and 132, notes on book 3, 15 and 31; see Schlemmer 1983a: 50–51). On the lexical level the Germanic impact was, according to Castelvetro (1714 [1572]: I, 35), limited to a few Langobardian cultural loanwords (cf. Melzi 1966: 166–167).

5.4.3 Cittadini 1601

The Italian philologist Celso Cittadini (1553–1627) discussed the same topics as Castelvetro: the sociolinguistic situation of the Roman Republic and Empire, and the transformation of Vulgar Latin into Italian. The sources of Cittadini’s linguistic thought on these topics are two treatises: Trattato della vera origine e del processo e nome della nostra lingua and Trattato de gli articoli, both published in Venice in 1601. ^[38]

Continuing Castelvetro’s criticism of Bembo, Cittadini proposed a new theoretical model, namely to apply the third Aristotelian category of change, i.e. alteration, to linguistic theory. Cittadini argues that Vulgar Latin changed into Italian by a gradual process of accidental changes in grammar. The categories of substance and accidents are crucial to the alteration theory. The substance is identified with the lexicon and the accidents with grammar. The Italian vernacular is the same as Vulgar Latin in its stock of words, and has only been varied in its grammatical accidents, without the destruction of the lexical substance. Cittadini refers to Aristotle as “il Filosopho” (Cittadini 1601: 38r, 40r; cf. Schunck 2003: 96).

As intimated in Section 1.5, Cittadini’s dependence on earlier sixteenth-century Italian linguistic theorists, such as Cl. Tolomei and L. Castelvetro, has been a subject of controversy. ^[39] Tavoni claims that “the role of Cittadini, who was formerly credited with postulating the idea of ‘vulgar Latin’, has been drastically reduced to utilizing, if not plagiarizing, both” (Tavoni 1998: 48). ^[40]

This statement requires clarification. In comparison with Tolomei, Cittadini is innovative in his hypothesis on Vulgar Latin. ^[41] Tolomei’s adherence to the corruption-generation theory prevented him from arriving at the concept of Vulgar Latin. His phonetic and morphological observations on Tuscan were certainly utilized by Cittadini in Le origini della volgar Toscana (1604), though not in Trattato della vera origine (1601).

In relation to Castelvetro, Cittadini made a number of improvements: he applied Isidore’s periodization of ancient Latin to the diglossia theory, discussed literary evidence relating to Vulgar Latin, provided epigraphic documentation for spelling and sound changes being characteristic of Vulgar Latin (Ward 1990: 105–106), and developed a new theoretical framework for analyzing the transformational process from Vulgar Latin to Italian, viz. the “alteratio” theory, although Castelvetro was also on this track (Faithfull 1962: 249–250).

To begin with, Cittadini (1601: 1r) opposed the corruption hypothesis, by which Italian is derived from the corruption of classical Latin at the time of the Goths and the Langobards. Cittadini (1601: 1v) used Isidore’s periodization of Latin into prisca, Latina, and Romana, mixta as the chronological context for deriving the Italian vernacular (“la lingua vulgare, nostra lingua”) from Vulgar Latin. He identifies three internal stages of the Roman period (lingua Romana), (i) until the time of Ennius and Plautus, (ii) until the time of Caecilius and Terence, and (iii) until the time of Vergil and Livy (Cittadini 1601: 4r), and tries to elucidate their phonological and grammatical features from inscriptional and literary sources (Cittadini 1601: 7v–33r).

Cittadini adopted Castelvetro’s diglossia theory of a social stratification of ancient Latin, namely with vernacular Latin as the low variety and cultivated Latin as the high variety, and also integrated it into his periodization of Latin.

In every period, both earlier and later, there have been two types of Latin in Rome: one rough and half barbarous, which was found among the populace, that is the unlearned Romans and foreigners, or I would say people of lower classes and rural population without learning; (Cittadini 1601: 2r)

[and] another variety cultivated by art and correct Latin, which was found among the noble writers and speakers and among educated people. (Cittadini 1601: 2 v)

The bifurcation of ancient Latin into the diglossic co-existence of two synchronic varieties is repeated many times. ^[42]

Paraphrasing and quoting Castelvetro, Cittadini (1601: 2v) argues that the ancient Vulgar Latin “la lingua latina antiqua del vulgo” has been preserved by the Italians up to his time and that its ancient name, vulgare, remains the most appropriate term for it. Cittadini (1601: 55r–57r, 61v–62r) is convinced of the terminological continuity in designating the vernacular as Latin and the Italians as Latins, for instance in Dante, Petrarca, and Boccaccio. He refers to latino volgare in Boccaccio’s Teseida (Cittadini 1601: 62r, 64r).

On the basis of Cicero (Brutus 258), Cittadini (1601: 34v) assumes that Latin deteriorated under the late Republic not only among low-class people (“la gente bassa”), but also among the writers (“gli Scrittori”). From his time, Cittadini argues “it was necessary to study the rules of Latin grammar or the works of ancient good writers of Latin in order to be able to speak well and purely (bene e puramente), i.e. Latin (cioè latinamente)” (Cittadini 1601: 35r). He continues: “Those who did not study were not able to speak Latin, but spoke with barbarisms, i.e. with vulgarisms” (Cittadini 1601: 35r). Cittadini here equates the high variety with good and pure language, and the low variety with barbarous language. The same equations also occur in another passage (Cittadini 1601: 44r, cf. 36r), where the low variety is viewed as lacking grammar (“senza grammatica”). Vulgar Latin is associated with an internal barbarization. For this reason, Cittadini describes Vulgar Latin as the younger sister of latina grammaticale (Cittadini 1601: 63r).

Cittadini makes a terminological distinction between ancient Vulgar Latin ^[43] and modern Vulgar Latin. ^[44] Cittadini argues that Vulgar Latin is the best term for referring to the Italian standard language of his time. He rejects the alternative terms Tuscan, Florentine, Cortigiana, and Italian (Cittadini 1601: 60r; cf. Ward 1990: 107–108).