The notion of language deviations in St. Augustine’s Ars pro fratrum mediocritate breuiata

Fábio Fortes

Universidade Federal de Juiz de Fora, Brazil

Fernando Adão de Sá Freitas

Universidade Federal de Juiz de Fora, Brazil

The idea that the grammarian was a kind of guardian of language, who should protect language against the terrible vices of barbarism and solecism was implicit in the old reflection on Latinitas, a concept derived from Rhetoric, but since the first century BC, attributed to the competence of Grammar.¹ In fact, passages in Cicero already attested that the discourse of Rhetoric considered the examination of Latinitas as part of a more elementary study, proper to grammatical formation.² It was for that reason, among others, that in the period of greatest flowering of Latin grammatical arts (between the third and fifth centuries AD), the concern with something like ‘linguistic correctness’ remained one of the most important pillars of this art, domain which contemporary commentators would refer to as the “third part” of the ancient grammars (Baratin & Desbordes, 1987, p. 215).

Latinitas represented both the model and the major criterion of what could be conceived as acceptable and unacceptable language expressions. If they diverged from this standard of reference, words and constructions could exemplify the so-called ‘vices’ - barbarism and solecism. Because they could make language less authentic, less 'Roman' in ancient terms, such impurities should be expelled from the language. However, for such grammatical deviations the Latin grammarians understood something very far from what we would call a ‘grammatical error’ today. In fact, the concern with language correctness did not represent an appraisal of what would be admitted as absolutely ‘correct’ or ‘wrong’ in spoken or written discourse, but stood for a range of criteria of adequacy for the language usage, defining what would be appropriate or inappropriate within some especific contexts of language production and reception, mostly in literary and rhetorical environments (Desbordes, 2007, p.97; Fortes, 2012).

In this article, taking as reference the grammatical treatise attributed to the late ancient philosopher Augustine of Hippo (c. 354-430 AD), the Ars pro fratrum mediocritate breuiata, we aim at describing the way Augustine dealt with the notion of ‘language deviation’, here understood as the phenomena of barbarismus and soloecismus. To this end, we seek to understand the philosopher's particular treatment given to the subject, relating it to the way barbarism and solecism are dealt with both in his ars grammatica and in some of his most important philosophical works, namely the Confessions, De ordine, and De doctrina Christiana.

We propose that a more strictly normative grammatical approach, connected to the defense of linguistic correctness (Latinitas), should, however, be relativised when put in perspective with the grammatical reflections present in the other non-grammatical works of Augustine. For this, our text is divided in two parts: in the first one we address some issues related to the text of the Ars breuiata itself and to the subject of linguistic correctness (Latinitas); then we try to show how the notion of language deviation in the Ars breuiata (barbarismus and soloecismus) must be understood in relation to the way in which these themes are treated in the Confessions, De ordine, and in the De doctrina Christiana.

The Ars pro fratrum mediocritate breuiata (henceforth Ars breuiata), is certainly among the least well-known works attributed to the Latin philosopher Augustine of Hippo (c. 354-430 AD). It is so not only because its very particular subject, quite different from the great theological and philosophical issues debated in his major oeuvre, but also due to philological reasons, which for a long time cast doubt on its authenticity or at least made scholars maintain some sceptical distance from this text (Law, 1984; Luhtala, 2005; Bonnet, 2013; Freitas, 2016).

A self-testimony about this text could be seen in the Retractationes (1.6), in which Augustine alluded to his writing of some works concerning the arts, specifically mentioning the conclusion of a work on grammar (de Grammatica).³ Notwithstanding this testimony, Augustine’s grammatical text has not been recognised for a long time, being quoted or even mentioned only by some grammarians of the mediaeval period, such as Cassiodorus (6th century AD) and Abbon of Fleury (1004).⁴ It was only after the nineteenth century, with the discovery of the manuscript V by Faustino Arevalo, that some attention was given to this text. In 1839, Cardinal Mai mentioned an Ars grammatica Sancti Augustini adbreuiata (Bonnet, 2013, p. viii). Weber's edition, with the full text, was published shortly after, in 1861. H. Keil, in his monumental edition of the Grammatici Latini (Vol. 5, 1868), did not adopt the full text of Weber’s edition, even though having this edition as his most important reference. Nowadays, after the study of V. Law (1984) and the critical edition of G. Bonnet (2013), there remains little doubt on the authenticity of the text.

In the domain of the Latin grammatical genre, the study of ‘language deviations’ (uitia orationis) was generally placed after the study of the parts of the sentence (partes orationis). Donatus’ Ars maior is by many scholars considered the most prototypical model of such organisation. In Donatus’ work one could clearly observe a threefold division, devoted respectively to the study of the littera, to the partes orationis and to the uitia and uirtutes orationis.⁵ Particularly, the third part would be the one in which the theme of Latinitas, concerning the vices and qualities of language, would then be developed: there would be a treatment offered to barbarism and solecism, among the vices, as well as to the metaplasm, the figures and the tropes, among the qualities.⁶

Augustine’s Ars breuiata also devotes his last part to the description and exemplification of the phenomena concerned with language correctness, but so it does exclusively with regard to solecism (soloecismus) and barbarism (barbarismus), not dealing with the so-called virtues of language, as did Donatus’ Ars maior. However, unlike the work of Donatus, Augustine provides us with a full definition of Latinitas (‘language correctness’), as we see in the comparison between the thematic structure of the Ars breuiata and the Ars maior:

Table 1:
Ars breuiata vs. Ars maior

The fact that Augustine’s ars has been introduced by the definition of Latinitas and not by a reflection on the minimal sound elements of the Latin language (the ‘Littera theory’), highlights the importance of language correctness (Latinitas) within this work. Such importance is also underlined by the fact that the language deviations (the phenomena of barbarism and solecism) are also addressed in other non-grammatical works by Augustine: in the De ordine, in the De doctrina Christiana, and in the Confessions, as shown in the table below.

Table 2:
Barbarism and Solecism in the works of Augustine

The very notion of Latinitas (here roughly translated as “language norm” or “language correctness”) is associated in Augustine to the way in which someone expresses him/herself through speech (loquendi), without producing any language corruption (incorrupte). For this, such standard of speech could only be attained by considering three criteria: the language logical system (ratio), authority (auctoritas) and usage (consuetudo):

Latinitas is the observance of correct speech in accordance with the language of Rome. It consists in three aspects, that is, ratio, auctoritas, consuetudo: ratio in accordance with ars, auctoritas in accordance with the writings of those who possess authority, and consuetudo in accordance with those things which are sanctioned and adopted by usage. (Aug. Ars breu. 1.1)⁷

The definition of Augustine is not exactly original: it repeats, for instance, the criteria already mentioned by Quintilian about the linguistic norm associated with oratory discourse: “There are special rules which must be observed both by speaker and writers. Language is based on reason, antiquity, authority and usage” (Quint. Inst. or. 1.6.1),⁸ being also repeated in the definition given by Diomedes (GL 1.439.10).⁹ Thus, most probably assuming a formulaic structure, akin to the Latin grammatical genre, Augustine’s definition comprehends both the adherence to a language standard (determined by auctoritas, ratio and consuetudo), and the necessity of performing language ‘Romanity’ - i.e. an ‘ethnic’ and ‘cultural purity’ of linguistic use, rather than a ‘stylistic purity’ or even less ‘an absolute correctness’ of language. That is to say, more than a style of prestige, a question of cultural identity seems to be here at stake: linguistic corruption denotes an equivalent impurity in the speaker’s origin or education.

In dealing with the theme of linguistic correctness, the curious thing is that whereas Donatus - and most of the authors of the arts¹⁰ - found it important to identify, describe and exemplify not only the language deviations (the barbarism and the solecism) but also the so-called ‘virtues of language’, Augustine dealt only with the deviant constructions, keeping a silence on the qualities of speech (uirtutes). And what would be the reason why Augustin has not given a word on the language virtues, limiting himself to discuss the vices? In order to tackle this question, it seems essential to examine the relationships the notions of ‘language deviations’ hold in Augustine's works as a whole.

As we have seen in the last section, barbarism and solecism were notions discussed not only in Augustine’s grammatical treatise, but also in other philosophical works. In this section, we intend to show the contexts in which such notions appear, in order to understand their role in the Ars breuiata.

In De ordine, when revealing to his mother, Monica, how such concepts could be hard to master, Augustine reports that:

Were I to say that you would easily attain a speech free from errors of grammar and pronunciation, I would lie through my teeth. I had to learn these things out of professional need, yet the Italians still correct my pronunciation of many words, and I in correct theirs. It is one thing to be certain in theory, another in practice with people. It is perfectly possible that an expert, upon examination, should find blunders in my speech [solecism]. There was indeed one who argued with me most convincingly that the great Cicero himself had made quite a few such blunders. So many alien words [barbarism] have been recently introduced in the language that even his famous speech by which he saved Rome would seem foreign today. (Aug. De ord. 2.17.45)¹¹

In this passage, Augustine identifies two types of language deviations: solecism and barbarism. His account highlights that, even having thoroughly studied such subjects, he was still reproached for his pronunciation. However, Augustine ponders that one thing would be to speak according to the precepts of grammar and rhetoric, another thing would be to express himself as the way common people (gente) did. This distinction made by Augustine between a speech organised according to grammatical and rhetorical precepts and a common speech seems to relativise some exaggeration related to linguistic purity, as the ironic comment at the end of the passage reinforces: “So many alien words [barbarism] have been recently introduced in the language that even his [Cicero’s] famous speech by which he saved Rome would seem foreign today.”¹² The mention of Cicero also brings to the Augustinian perspective the character of authority attributed to the classical Latin orator (Law, 1987; Cameron, 1993).¹³

In the Confessions, the notions of solecism and barbarism are also related to an ethical discussion: the difference between the purity of language with respect to form (i.e. when deprived of solecisms or barbarisms), and purity concerned with content (i.e. language conveying either moral or immoral themes):

But what wonder was it, if I were thus carried towards vanity, and estranged from thee, O my God; whenas such men were propounded to me to imitate, who should the deliver any of their own acts, though not evil, with barbarism or solecism, they were utterly dashed out of countenance: but should they make a copious and neat oration of their own lusts, in a round and well followed style, would take a pride to be applauded for it. (Aug. Conf. 1.18.28)¹⁴

Just as presented in De ordine, the remarks on the barbarism and on the solecism in the Confessions would relativise the zeal of those who were excessively concerned with the purity of language. As we can see from the except above, more important than having a richly ornated and elegant speech (ornate copioseque), someone would better care for the ethical dimension of his/her discourse - that is to say, if it conveys either noble and virtuous or shameful and dishonourable content. Therefore, any language deviation, either a barbarism or a solecism, would be less shameful than a moral deviation. Augustine's comment, therefore, is in the same sense as the one we observed in De ordine above: it is less important to master the virtue of language (in the plane of form) than the virtue in language (in the plane of substance). In any case, both passages do no more than commenting on the phenomena of barbarism and solecism, they do not provide us with any definition of them. A clear definition of them could only be found in the Ars breuiata and in the De doctrina Christiana, as follows:

Solecism is the vice of the language that is made through Latin [words], in certain parts of the sentence, when they are linked together. Whoever, in effect, says inter hominibus, if we considered every word isolated, he would not have commited a vice, for both inter and hominibus are Latin words, but, if united to one another, it represents a deviation. Therefore, in the same way, when an error is made in any logical part of the sentence, which was related to the eight parts of the sentence, it is called solecism. (Aug. Ars breu. 1.98)¹⁵

What is called a solecism is simply what results when words are not combined according to the rules by which our predecessors, who spoke with some authority, combined them. Whether you say inter homines or inter hominibus does not matter to a student intent upon things. (Aug. De doc. Chr. 2.44)¹⁶

In the Ars Breuiata, Augustine characterises solecism as the violation of the logical organisation of a sentence (ratio). The example given by Augustine clearly shows this problem: the preposition inter would require, according to the rule, the accusative case, not the ablative, as in inter hominibus (ablat.). It is implied therefore the existence of an alternative and regular construction, with accusative (inter homines).

In the excerpt from the De doctrina Christiana, it is made explicit that the norm of language would be the accordance with the authority of predecessors (auctoritas priorum), although the philosopher defines solecism in the same way, that is, as a deviation related to the construction of the words in a sentence (also giving the same example). Although alluding to the same phenomenon and defining it in a quite similar way, these two passages focuses on two of the three criteria by which Latinitas had been previously defined: the logical organisation of language (ratio) and the authority (auctoritas). Moreover, the two definitions present specificities regarding the genre of the texts in which they are found. In the Ars Breuiata, a technical treatise on grammar, Augustine emphatically states that the phrase inter hominibus would be, in fact, a vice associated with elocution (uitium locutionis). Augustine uses the verb peccatur (to sin / to fail) to indicate this deviation from the norm. In addition, Augustine mentions that solecism occurs when the parts of speech are ordered in a way that disobeys to that rule responsible for the regular and natural concatenation of sentences in the language (ratio).

In the De doctrina Christiana, however, even though it displays some comments proper to grammatical discourse, such reflections seem to be of secondary importance in relation to a more advanced philosophical discussion. It means that a minor deviation in the rules of grammar (for instance: saying inter hominibus instead of inter homines) would be of less importance for those “seeking the knowledge of things” (ad rerum non pertinet cognitorem). It means that, likewise we have seen in the Confessions and in the De ordine, there seems to be some relativisation of the normative concern, in favour of a more philosophical pretension, as we can see again in the Confessions:

Behold, O Lord God, and patiently behold, as thou still dost, how diligently the sons of men observe the rules of letters and syllables received from former speakers; and yet regard not the eternal covenants of everlasting salvations, received from thyself. Insomuch, that he who either holds or teaches the ancient rules of pronunciation, if contrary to grammar he shall pronounce ominem, (that is, a man) without H in the first syllable; he shall displease men more, than if against thy rules he should hate a man, although he be a man. As if any man should think his enemy to be more pernicious to him, than that hatred of his own is, whereby he is set against him: or imagine that he does worse scath to another man by persecuting him, than he does to his own heart, by contriving enmity against him. Certainly there is no more inward knowledge of Letters than this law of conscience, that one is doing to another what himself would not suffer. (Aug. Conf. 1.18.29)¹⁷

Augustine does not ignore the relative importance of grammar knowledge for one’s education (the fact of dealing with such grammar facts in so many books seems to confirm this). However, he seems to consider such knowledge not as an absolute norm for language or the very end of one’s studies: he rather puts ‘language correctness’ in the perspective of a ‘moral correctness’. That seems to be the reason why Augustine is much concerned to deal with the ‘language deviations’ in his grammar manual, and not with the ‘language virtues’ (the ornaments of language): the importance lies on tracing a parallel between the vices in language and the vices in life (his goal is not offering something like a ‘manual of style’).

On the barbarism, the definitions given are the following:

Likewise, what is a barbarism but a word articulated with letters or sounds that are not the same as those with which it was normally articulated by those who spoke Latin before us? Whether one says ignoscere with a long or short third syllable is of little concern to someone beseeching God to forgive his sins, however he may have managed to utter the word. What, then, is correctness of speech but the maintenance of the practice of other’s, as established by the authority of ancient speakers? (Aug. De doc. Chris. 2.45)¹⁸

And in the Ars breuiata:

There is barbarism when, individually, the same words are not Latin, as if one says hominem, without aspiration, he commits barbarism, or when saying coronam someone adds aspiration; when a syllable is withdrawn in luctat, naturally it will be a mistake; or if to potest another syllable is added like potestur, it will not be Latin. Now if saying pone the first syllable is abbreviated with the removal of time, there is barbarism. Now if by saying bonus, the first syllable is allonged with the addition of time, there is also a mistake; or if the name is pronounced and the syllable is not accentuated, bare for removing the accent; if, however, the two syllables are accentuated in a part of the sentence, by the addition of the acute accent, I disturb the ear; if we say uulla instead of uilla, with the change of the letters, if we say displicina instead of disciplina, with the change of the syllable, there is a mistake. The removal, addition, change, alteration, or the aspiration of letters as well as syllables or accents or, sometimes, times, constitutes a barbarism. (Aug. Ars breu. 1.100)¹⁹

The approach given to barbarism in both contexts are quite similar. In both of them, Augustine refers to words which would have been perfectly Latin words, had not been corrupted with respect to his pronunciation or writing. While in the De doctrina Christiana we are given a definition for the phenomenon with few examples, in the Ars breuiata we are shown many examples: either from the point of view of the pronunciation of phonemes (e.g., saying hominem without the initial aspiration [h], or adding an unnecessary aspiration to the word corona, pronouncing it chorona); either from the point of view of their morphology (the addition, alternation or subtraction of syllables, as in saying potestur instead of potest, or displicina instead of disciplina etc.), or from the point of view of syllable quantity (as in saying pone with its first [o] brief and not long as it should be).

Without considering, for while, each language occurrence in these examples (which would deserve a study in itself, since it reveals the dynamics of linguistic change back at Augustine’s times), it is important to note that the definition given in the first excerpt highlights the third defining criterion of Latinitas: usage (consuetudo). Again, while we see a more linguistic oriented approach in the Ars breuiata, we are given further remarks in the philosophical treatise, in the sense of minimising the excessive importance someone could give to the notions of language error. In fact, Augustine states that for whoever is begging God for forgiveness, it would be irrelevant to pronounce the word ignoscere (‘to forgive’) with a long or brief syllable before last. In sum, the strictly normative approach, apparently present in the Ars should therefore again be softened pro bono, for a more far reaching reason, now of a philosophical order, since language is committed to the primacy of Christian thought.

As far as the treatment given to the phenomena of language correctness (Latinitas) goes, Augustine reflections could be summarised as follows:

1. 1. Augustine’s grammatical treatise, the Ars breuiata, is limited to dealing with the language deviations (uitia orationis): the barbarism (barbarismus) and solecism (soloecismus). Differently from what used to be common in the Latin artes grammaticae (sc. in Donatus’ Ars maior, for instance), it does not give any attention to the ‘language virtues’ (uirtutes orationis);
2. 2. As far as language deviations are concerned, while the treatment of barbarism and solecism is quite practical in the Ars breuiata - with plain definitions followed by many examples -, Augustine presents further critical remarks on this issue in his philosophical works - in the Confessions, De ordine and De doctrina Christiana. In these last three books, the strictly linguistic and normative conception associated with barbarism and solecism in his grammatical work is replaced by a relativised concern towards language: Augustine emphasises that a language correctness could not be more important than moral correctness.
3. 3. The three criteria by which Augustine defined language correctness in his Ars breuiata (namely ratio, auctoritas and consuetudo) is also mentioned in his three philosophical works in which the reflection on solecism and barbarism also raises.

With no pretension of advancing a final conclusion to this study, we suggest that the treatment given to language deviations in the Ars breuiata should not be read without comparison to what is presented in those philosophical works. Because the Ars breuiata seems to play the role of a more propaedeutic reflection on language, it must have been envisaged to be a more basic and brief (breuiata) language reference to Augustine’s later philosophical reflections carried out in his work of maturity.