SASSOFERRATO PAPER 2003

This paper falls into three parts. First, I review the basic facts relating to Perotti's Rudimenta grammatices. Second, I describe the overall structure of the work and in particular the treatise on letter writing, De componendis epistolis, that forms the third and final part of the book. Finally, I examine the little evidence known to me regarding possible predecessors who may have influenced Perotti in composing his manual of Latin epistolary composition.

From internal evidence, we know that Niccolò Perotti (1429/30 to 1480) completed the Rudimenta grammatices towards the end of 1468. At that time, he was in his upper thirties and living in Viterbo, where he was papal administrator. A manuscript of the Rudimenta grammatices has been preserved and is now in the Vatican Library (shelf mark: Vatic. Lat. 6737). This manuscript was identified as an autograph by Cardinal Giovanni Mercati in his seminal 1925 monograph on Perotti's life and writings. (For a complete reference, see the bibliographical appendix below.)

The first printed edition of the Rudimenta grammatices appeared on 19th March 1473, hence only a little over four years after Perotti completed the manuscript. The printers of that edition were none other than Conrad Sweynheym and Arnold Pannartz, Rome's first printers, two key figures in the history of printing in Italy. A spate of printed editions of the Rudimenta appeared after that first edition of 1473. Wolfgang Milde, a manuscript librarian at the Herzog August Bibliothek in Wolfenbüttel, counted 176 editions (see his 1982 article), but there are others that he was unaware of. (One of the more notable ones is an edition printed in Venice in 1474 of which the only known copy is in the library of the Russian Academy of Sciences in Saint Petersburg; see the book by Larissa Stepanova listed in the Bibliographical Appendix below.)

The historical importance of Perotti's Rudimenta grammatices lies in the fact that it is the first complete grammar and stylistics of Latin to be produced in the Renaissance period. Earlier humanists, like their immediate predecessors in the Trecento, composed only brief grammatical manuals for their students. The earliest example of that genre of grammatical writing in the Quattrocento is a work usually titled Regulae grammaticales that was composed by the great humanist Guarino Veronese (1374-1460) sometime in the second decade of the fifteenth century (it is known to have been in circulation by 1418). However, manuals of this kind had limited coverage: essentially they were designed to supplement standard textbooks like Priscian's Institutiones, not to replace them. By the middle of the Quattrocento, therefore, there still existed no single grammatical manual that could be used at all levels of the grammatical curriculum. It was this lacuna that Perotti set out to fill in the late 1460s, and we may speculate that it was precisely because the Rudimenta grammatices filled this lacuna that it immediately became a best seller and remained so for several decades. Thus, Perotti's was the first omnibus grammar of Latin.

As regards the dimensions of the work, the autograph manuscript of the Rudimenta grammatices preserved in the Vatican Library contains 143 leaves, hence 286 pages. Each page has thirty-three lines, with an average of eight words on each line. That yields a total of over seventy-five thousand words (33 x 8 x 286 = 75,504). The first printed edition, on the other hand, has 114 leaves, of which the first and last are blank, hence 224 pages of text. There are 38 lines on each page, hence 8,512 lines in all. If we assume that there are about nine words on each line, that yields a figure of 76,608 words for the entire text. If we average the figures for the manuscript and the printed edition, we get 76,056. It is therefore a work of considerable proportions.

The Rudimenta falls into three clearly defined parts. The first is a beginner's manual similar to the Ars minor of Donatus and other works of that type that were popular throughout the Middle Ages. The second section of the Rudimenta, headed "De constructione orationis," covers a variety of intermediate-level topics, such as sentence construction, gerunds and supines, relative pronouns, patronymics, the comparison of adjectives, and figures of speech: these are the topics that Guarino Veronese covered in his Regulae grammaticales. Finally, the third part is a manual of epistolary style and is headed "De componendis epistolis."

The entire Rudimenta grammatices is composed in the humanistic style exemplified by the Regulae grammaticales of Guarino Veronese. It was the third part, the manual of epistolary style that made the Rudimenta a comprehensive grammar and stylistics of Latin. One should recall that in the Quattrocento successful Latin composition in the classical style was the ultimate goal of grammatical instruction. Hence, if one added a manual of letter writing, the result was something that covered more ground than any single grammatical manual that had been available hitherto.

The Rudimenta grammatices proceeds throughout in question-and-answer style, like a catechism. Thus, the first paragraph of the treatise on epistolary composition reads: "Quare inuentae sunt epistolae? Vt eos cum quibus siue propter absentiam, siue propter ruborem, seu ob aliam quamuis causam loqui non licet, certiores facere possimus si quid sit quod eos scire oporteat, siue nostra siue illorum siue aliorum causa" (Vatic. Lat. 6737, f. 96v).

Religious doctrine had, of course, long been taught in the form of a catechism (and to some extent still is to this day). From late Antiquity on, however, this method was also widely used by teachers composing elementary manuals of various kinds. For example, the Ars minor of Donatus, which dates from the fourth century A.D., was in this format, and so was another grammatical manual popular in Italy during the late Middle Ages that scholars now call the Ianua. However, both the questions and the answers in the Ars minor and the Ianua are shorter than the pair of sentences from Perotti's Rudimenta that I have just quoted.

The second major part of the Rudimenta grammatices was based on earlier humanistic manuals, such as the one by Guarino Veronese just mentioned. When I say based, however, I am not implying that Perotti copied verbatim, but merely pointing out that the format and methodology were the same. In reality, the relationship between the two works is complex.

Perotti's treatise on letter writing, the third part of the Rudimenta grammatices, is not obviously based on any single earlier humanistic work. It contains a number of ingredients woven together. Here I shall have time to discuss three of these: (1) instructions on how to compose a letter, (2) discussions on how to distinguish the meanings of easily confused words and avoid non-classical words and turns of phrase, and (3) more general lexicographical and etymological information. None of these ingredients was new. Manuals of written composition had been popular in the Middle Ages. This branch of instruction was called dictamen, and manuals of this kind were referred to as artes dictandi. This genre flourished in northern Italy from about 1100 onwards: it played an important role in the education of notaries and other legal specialists.

This epistolary component of the "De componendis epistolis" is arranged as follows. Perotti imagines his student composing a personal letter to a friend, hence not a business letter of the kind that had been the focus of attention in the earlier artes dictandi. Perotti's student first formulates what he wants to say in his vernacular, and then possible Latin equivalents of these pieces of vernacular are exhibited. The imaginary letter begins as follows: "Quomodo eleganter dicemus Io ho receuuta la tua littera? Multis id modis dici potest: Accepi litteras tuas. Accepi abste litteras. Redditae mihi fuerunt litterae tuae. Redditae mihi fuerunt abste litterae. Reddidit mihi Helius Perottus litteras tuas. Accepi quas ad me Helio Perotto dedisti litteras. Accepi epistolam quam ad me scripsisti. É Thusculano reddita mihi est epistola tua. Attulit mihi Helius Perottus epistolam tuam" (Vatic. Lat. 6737, f. 99r). At the end of the treatise there is a complete model answer, i.e. a translation of all the vernacular sentences composing the imaginary letter in the order in which they were discussed.

The second main ingredient, the correction of non-classical syntax and the focus on the correct choice of words, is interspersed throughout: it is hung on the epistolary thread, even though this thread is sometimes tenuous. There is no alphabetical index of words covered. A student or instructor consulting Perotti's grammar could not look up information on a particular word without knowing where to find it. It may be recalled that printed editions of Perotti's Cornu copiae are provided with two alphabetical lists, one for the Latin words and the other for the Greek words. Similarly, printed editions of Lorenzo Valla's Elegantiae also contain an alphabetical list of words. In contrast, there are no indexes of any kind in the Rudimenta grammatices, and in the manual of epistolary composition there are not even subtitles to guide the reader. The heading that introduces the manual of epistolary style is an ordinary centred heading, no different graphically from many other centred headings that introduce relatively unimportant subsections in previous parts of the Rudimenta grammatices.

It is interesting to note that Perotti placed in the margin of his autograph manuscript key words referring to matters covered at those points in the main body of the text, but the printers ignored these marginalia. (Other marginalia of Perotti's, on the other hand, are incorporated in the text.) Moreover, paragraphs in the manual on epistolary style are much longer than those in the rest of the Rudimenta, and the digressions to discuss the differences between words are at times long and involved. All this means that the user could more easily lose his way there than elsewhere in the Rudimenta. I realize, of course, that the indexes in Perotti's Cornu copiae and Valla's Elegantiae may not have been compiled by their respective authors when they were being composed. However, it is significant that they made their appearance in printed editions at all. In contrast, no printed edition of the Rudimenta grammatices that I have so far seen comes equipped with an index.

A curious feature of the epistolary component in the "De componendis epistolis," is that of the five traditional parts of the letter, viz. salutatio, exordium, narratio, petitio, and conclusio, Perotti treats only the first two and in effect equates salutatio and exordium. The words narratio, petitio, and conclusio do not occur anywhere in the third part of the Rudimenta, where one imagines that they might belong. On the other hand, if we consult, say, the encyclopedic Catholicon of John of Genoa, dating from 1286, we find extensive articles on narratio and petitio, especially the latter. It is also interesting to note, for comparison, that in the "De conficiendis epistolis libellus" attributed (falsely, it would appear) to Lorenzo Valla in the Opuscula quaedam nuper in lucem edita (Venice: Cristoforo de Pensis, 1503; pp. 97ff), all five traditional parts of the letter are handled. One assumes that Perotti's non-traditional approach to the parts of the letter was intentional.

Therefore, while one presumes that the addition to the Rudimenta grammatices of an entire manual dealing with epistolary composition was an attempt on Perotti's part to bring the work in line with contemporary tastes, it is significant that he steps out of line in certain ways, e.g. in omitting any serious discussion of how to plan the layout of a letter or how to adapt a letter to accord with the relative social status of the intended recipient. Perotti narrowly focuses on intimate letters exchanged between social equals. His model is Cicero's Epistulae ad Familiares, not the business letters that the dictatores of the previous age had trained their students to write. To that extent, one suspects that Perotti's manual of epistolary style may have been less useful, in practical terms, than those of his Trecento predecessors.

Finally, instead of a conclusio, Perotti substitutes a brief peroratio, addressed to the dedicatee of the entire Rudimenta, his nephew Pirro Perotti. Perotti's peroration is in fact more like a conclusion than a Ciceronian peroration. Many of Cicero's perorations are, of course, quite lengthy. This may be yet another departure from tradition on Perotti's part.

The second major strand in Perotti's manual, the correction of non-classical vocabulary and syntax, has obvious affinities with Lorenzo Valla's Elegantiae, a work of which Perotti was well aware. Here is one example of that feature of the work: "Hoc loco notandum quod a uirtute non dicitur uirtuosus, quo uerbo imperiti utuntur, quemadmodum etiam Graeci apo tes aretes non faciunt adiectiuum, sed cum uirtute praeditum dicere uolunt, dicunt spoudaios, hoc est studiosum, quod Cicero imitatur multis in locis, ut in libris ad Herennium" (Vatic. Lat. 6737, f. 132v). This may be compared with Lorenzo Valla, Elegantiae, book 1, chap. 21 headed "De nominibus in osus," where the Latin word uirtuosus and Greek spoudaios are mentioned. I could multiply examples of this kind many times.

Not surprisingly, Perotti's lexicographical and etymological urge is much in evidence in his "De componendis epistolis." This is the same sort of material with which he later filled the Cornu copiae, and it is inserted in the discussion of other topics. Any word mentioned may set him off on a tangent. Here is a brief example of this pervasive feature: "Quae differentia est inter salubrem et sanum? Saluber siue salubris dicitur quod saluti conducit, ut aer, cibus, potus, exercitatio, et similia. Sanus uero cui salus inest, ut homo et animalia. Salubre est quod dat, sanum quod accipit salutem" (Vatic. Lat. 6737, f. 134r). Again, I could cite many more examples.

Finally, let me make a few preliminary remarks about how Perotti's treatise on letter writing is related to similar works produced by earlier humanists. (On the general topic of letter writing by humanists in the Quattrocento, see Cecil Clough's study cited in the Bibliographical Appendix.)

The earliest humanistic effort in this area appears to have been two works by the Paduan humanist Gasparino Barzizza (c. 1370 to 1431). One was entitled Suavissimum epistolarum opus or the like (see Hain 2668), and the other Exempla exordiorum (see Hain *2682, Gesamtkatalog der Wiegendrucke, no. 3690). The former was a collection of model letters, and the latter a collection of exordia only. (Mercer discusses them both in his monograph on Barzizza, cited in the Bibliographical Appendix.) Since I have so far seen neither of them I cannot yet comment on their possible relations with Perotti's "De componendis epistolis." A manual of epistolary style was also attributed traditionally to Lorenzo Valla, as I have already mentioned, but my understanding is that it is supposititious. In any case, there are no textual affinities between it and Perotti's manual of epistolography, as far as I can see.

A fifteenth-century humanist who definitely wrote a treatise on letter writing was the Sienese Agostino Dati (c. 1420 to 1478). He is thought to have studied with Francesco Filelfo. Dati's Elegantiolae most probably pre-dated Perotti's Rudimenta grammatices, but whether Perotti saw it before he composed his own "De componendis epistolis" is unclear. It is conceivable that he might have done so: the two works are similar in emphasizing stylistics and phraseology. After having examined two printed editions of the work, I conclude that while Dati certainly handles much the same material as Perotti there are no textual coincidences between Dati's and Perotti's treatises. Furthermore, one significant difference between them is that there are no model letters in Dati's treatise, as Cecil Clough points out (see p. 47 of his article cited in the Bibliographical Appendix below).

Another fifteenth-century humanist who wrote a manual of epistolary style was the Franciscan Lorenzo Gugliemo Traversagni of Savona (1425-1503). While a student in Vienna in the 1450s he composed a brief Modus epistolandi. Subsequently, a number of printed editions appeared, beginning in 1480, but it seems that they all came out in northern Europe. I have not yet seen a copy of this work, but since Perotti died in 1480 he could not possibly have been familiar with any of those printed editions. On the other hand, it is possible, though perhaps not probable, that he saw one of the manuscript versions of this work (see Ruysschaert's remarks on pages 200ff. of his article cited in the Appendix).

Summing up, it is certainly conceivable that Perotti was aware of one or more immediate predecessors in the fifteenth century who wrote treatises on letter writing, and he may well have modelled his "De componendis epistolis" on one of these earlier works, but I have so far been unable to identify any clear links. Further research on this question is under way. The relation between certain aspects of Perotti's "De componendis epistolis" and his later Cornu copiae is too obvious to require further elaboration.

In the form in which this paper was presented orally on 3 July 2003, I was not able to draw the attention of my audience to the voluminous literature dealing with Niccolò Perotti and related issues. In the eventual published version of this paper, I shall obviously follow the usual practice of directing the reader to as much of the relevant secondary literature as I can. What follows here is an attempt to list basic bibliographical items that are either mentioned in this document or that I consider especially relevant to the questions that I raise here. It does not claim to be a comprehensive bibliography.