The second part of the seminar “Attributive philology and digital analysis of literary texts”, organised by Humanities for Change, promoted by Venice Centre for Digital and Public Humanities (VeDPH) and financed with funds for student activities of Ca’ Foscari University of Venice,took place on Friday, December 11, 2020 from 10:00 a.m. to 12:00 a.m. and was opened by Irene Mamprin's welcome address, followed by an introduction based on some remarks by Pasquale Stoppelli collected in the essay Fare filologia nell’era digitale (in Elogio della lentezza. Lezioni Sapegno 2002, edited by Gian Luigi Beccaria, Torino, Aragno, 2002, pp. 35-45). The transition from paper to digital has significant repercussions on philological and literary research. The electronic text, much cheaper than a printed one, is manipulable and duplicable: a user can make changes through word processing programs and create a personal copy, just as the medieval amanuensis could manipulate and correct his codex. Moreover, duplicating a digital text has no cost, and is therefore easily shared on the network in an anonymous form; this practice tends to a weakening of those principles of authorship and intangibility of the text that have been established during Humanism, in particular thanks to the advent of printing. Nevertheless, the new information technologies can be of great support to the philologist, qualifying as a heuristic means of knowledge otherwise unattainable. Texts, as sequences of alphabetical symbols interspersed with spaces, can be processed by the computer through binary sequences; with a digitized text and a software capable of generating in a few seconds concordances, rhymes, statistical indices, incipitaries, and so on, it is possible to extract a quantity of information far superior to that of a manual search. This tool allows phono-morphological, syntactic, semantic and intertextual research; it also allows statistical, stylistic and usus scribendi surveys. Given the amount of data that can be deduced from this type of instrument, the warning is first of all to ask circumscribed and qualitative questions that move from a defined hypothesis; secondly, the researcher must carefully analyze the results: a fragmented text could provide a misleading vision and the presence of an element of style in two authors is not necessarily the sign of a filiation, because there could be intermediate texts not recorded in the archives or it could turn out to be a widely diffused element in a given genre or cultural context. This should not be translated into an element of distrust towards digital technology, but rather into an awareness of the potentialities and limits of the medium: no literary phenomenon is automatic and the evaluation of data always requires a critical eye, a judicium of the data found, which the philologist and the critic cannot avoid.
Paolo Mastandrea and some cases of poetic memory in Latin poetry
After Mamprin's introduction, Professor Paolo Mastandrea took the floor and presented some examples of poetic memory, that is, of unconscious phonic recoveries that are established between several texts. A fundamental role for the advancement of these studies is played by the Musisque Deoque database, which allows the search for metric-verbal occurrences.
”Intertextuality is not mere spasmodic exercise in order to recover as many links, quotations, allusions as possible: intertextuality is the history of the fortune (or misfortune) of texts.Paolo Mastandrea
With this statement the first speech of the day begins, followed by a reference to an example: although the text of Lucretius was not accessible before 1417, direct reminiscences by authors between the end of the fifth and the beginning of the sixth century fix the terms post quem the author would have disappeared to resurface then thanks to Poggio Bracciolini. Subsequently, through the use of Musisque Deoque, it was possible to identify not only clear resumptions but above all obscure and perhaps unconscious reminiscences. The first case concerns the analogies between a verse of an elegy by Propertius and another within a joking epigram in which the author, Ausonius, mocks a character because of his ugly voice. Between "flare, nec Aonium tingere Marte nemus" (Propertius, Elegiae, 3, 3, 42) and "cum uis Arcadicum fingere, Marce, pecus" (Ausonius, Epigrammata, 80, 4) parallels are identified that cannot be coincidences. Aonium is an adjective connected to the forest of Helicona, Arcadicum indicates Arcadia, the favorite place of the bucolic: the symmetry is established between different locations in Greece. Paying attention to the contexts in which this occurrence appears, it is clear that there is no semantic connection: on the one hand Propertius reports the words of the Muse, who suggests to the poet not to devote himself to epic but to the soft verses of elegy; on the other hand Ausonius compares the sound of the voice of a certain Mark to the braying of donkeys. The second part of the speech focused specifically on the relationship between Propertius and Ovid (who survived the first for about thirty years), of which are observed several textual convergences also in relation to other authors. A notable case concerns the equivalence between "Penelope poterat bis denos salua per annos" (Propertius, Elegiae, 2, 9, 3) and "Stat uetus et multos incaedua silua per annos" (Ovid, Amores, 3, 1, 1), but specifically between the former and "utque bibant Tyrium bis quinos saeua per annos" (Silius Italicus, Punica, 16, 215). In this case Propertius describes the situation of Penelope, kept pure for twenty years, waiting for her husband; Silio Italico, author of the late Flavian age, refers instead to the years of Hannibal's stay in Italy. In all of these examples, a change in the consonants is enough to create completely different verses (and contexts), substantially altering the meaning. No one among the commentators has ever been able to trace this type of phonic-rhythmic recovery because lexicological instruments are not able to replace letters with sounds. Musisque Deoque's metric-verbal co-occurrences function, one of the most sophisticated offered by the program, was able to detect similarities in vowel succession within seconds.
Pasquale Stoppelli and some Machiavellian discoveries
After Mastandrea's talk, Pasquale Stoppelli's speech focused on Niccolò Machiavelli, in particular on the Mandragola and on the novella of Belfagor (for which you can see the book Machiavelli e la novella di Belfagor. Saggio di filologia attributiva, Roma, Salerno Editrice, 2007). The tradition of the Mandragola relies mainly on two witnesses, a print and a manuscript, both dependent on a lost non-autograph antigraph and bearing a large number of adiaphora variants that make it difficult to reconstruct the text. Faced with such a compromised situation, it was possible to resolve some doubts thanks to the electronic tool. A first case concerns the contrast between the correlation “tale … quanto" of the manuscript and “tanto … quanto” of the print. The examination of the digital database has allowed us to identify the use of the second expression in several middle level Florentine authors between the end of the fourteenth and the beginning of the sixteenth century (Franco Sacchetti, Alessandra Macinghi Strozzi, Pievano Arlotto). This discovery allowed Stoppelli to conclude that “tale … quanto" is the variant to be put in the text as a lectio difficilior with respect to the banalized alternative. It was also doubtful the choice between the equivalent lessons “garzonaccio" (of the print) and “giovanaccio" (of the manuscript), an adjective that in the text refers to Callimaco himself. Digital research has allowed the editor to resolve this impasse: the second attribute, in fact, is found in Boccaccio's Decameron in reference to Masetto di Lamporecchio (III, 1), hired by the nuns as a vegetable gardener because of his (fake) muteness, so that he could not report what happened between himself and the nuns. The concordance within two analogous narrative situations led Stoppelli to the conclusion that “giovanaccio” is, therefore, the original lesson.
The digital research was not only useful in the choice between adiaphora lessons: it led to interesting reflections on the date of composition of the Mandragola that Roberto Ridolfi fixed at 1518, in his 1965 edition of the work, after writing for only two months. This identification was based on an indisputable fact, that is the presence, in the comedy, of phrases taken from the vulgarization of Andria in 1517. This last dating was based, in turn, on some studies on the graphic uses of Florentine. The research, however, had been conducted on a reduced corpus of manuscripts. For this reason, the work was carried out again by Stoppelli, following the same methodology but taking as reference the edition of all the works of Machiavelli by Martelli (published in 1971) studied in their original handwriting. The comparison allowed to establish that the graphic system of the codices of the text could not refer to the years 1516-1519; this element, together with the poor quality of the vulgarization and some particular renderings (for example the translation of "Edipus" with "prophet", a way in which Savonarola was ironically indicated - a mockery that the author would not have allowed himself to propose in the years 1516-1519) led Stoppelli to consider the text as an early work, to be placed between the years 1494-1495. As a counter proof, the edition of Andria of 1494 for Lazaro de' Soardi, unlike the following editions, does not present the text in acts, like the Machiavellian one. Once this obstacle has been removed, the editor can hypothesize a date of composition of the Mandragola between the years 1514-1520, a chronology that explains the quotation of phrases taken from Francesco Vettori's letters to Machiavelli in the same years.
The second part of the intervention dealt with the only novella recognized to the Florentine author, that is the Belfagor arcidiavolo of which we possess the autograph. The first edition of the text dates back to 1549, promoted by Machiavelli's son in response to a Venetian priest, Giovanni Brevio, who had published five years earlier a text coinciding for about 35%. There were three hypotheses: either Brevio had copied from Machiavelli, or vice versa, or both had referred to a third lost writing. From the digital analysis it emerges that the non-coincident part concerns Machiavelli's phrasing: if Brevio had plagiarized the work, then he should have had such a philological knowledge of the Florentine language that he should have purged the text of those typical stylistic features that only today, in part, we are able to recognize.
Marco Sartor opens the discussion with a question about the possible risk of an incorrect use of databases by inexperienced scholars, which could cause a tendentious drift such as the filling of comments with a rudis indigestaque moles of parallel passages, that is, intertextual references reported without a critical examination and deprived of the integration in a context that clarifies their relationships. Sartor asked the speakers what advice they could give to young scholars in order to reach a sort of a "middle ground" when attaching parallel passages. "Information technology - Mastandrea replies – has not made things worse: it does a "slave" job, leaving more time for the intelligence needed to discern the useful from the superfluous". The digital medium is, therefore, a tool that must facilitate the search for sources to leave more time for the philologist's judgment. "One is no longer scientific by amassing, but by selecting" Stoppelli immediately emphasizes, referring in particular to critical apparatuses rich in variants that are useless for the study of the text.
After a quick speech by Nicola Cittadini on the concrete research practice that allowed the tracing of cases of phonic-metrical memory on Musisque Deoque, Irene Mamprin raised the problem of the impediment constituted by copyright in the sharing of knowledge, the basis of progress, in relation to Italian literature of the twentieth century. Having strongly emphasized the legitimacy of copyright, Stoppelli hoped for the use of an adequate software that, when faced with the impossibility of reading a text in its entirety, could still highlight the concordances through a parcelling out of the work. Mastandrea's position is similar to his colleague’s. He is against complete gratuitousness, since freedom of expression is to some extent always linked to the economy.
Sartor's concluding question, which will be followed by final greetings and thanks, asks about the future perspectives in relation to potential Italian and European funding to invest in databases. They could be allocated to the creation of search engines able to retrieve phonic and metric-verbal occurrences or to extend the search to synonyms or between different linguistic codes. Stoppelli is cautious in hoping for subsidies on literary textuality stricto sensu; more confidence is placed in research in computational linguistics which deals with the control of language also for economic purposes. Paolo Mastandrea, on the other hand, is more optimistic, recalling the Venetian reality in which much is being done to continue and improve a work begun almost twenty-five years ago, as demonstrated by the presence of a center for Digital Humanities, the Venice Centre for Digital and Public Humanities, which counts among its members figures of great professionalism and competence. Finally, a project of national interest will be announced in the coming months and realized in close collaboration with the Istituto di Linguistica Computazionale “A. Zampolli” of the CNR, which has opened a branch within the Department of Humanities of Ca’ Foscari University of Venice.
To learn more
- Mastandrea, Paolo. 2017. «Sui principî della poesia, la ricerca intertestuale con strumenti elettronici». In Filologia digitale: problemi e prospettive, 73-111. Roma: Bardi.
- Stoppelli, Pasquale. 2002. «Fare filologia nell’era digitale». In Elogio della lentezza. Lezioni Sapegno 2002, a cura di Gian Luigi Beccaria, 35-45. Torino: Aragno.
- Stoppelli, Pasquale. 2007. Machiavelli e la novella di Belfagor. Saggio di filologia attributiva. Roma: Salerno Editrice.