Carolingian Critters IV: Leiden, Universiteitsbibliotheek, BPL 67F. A peep into the workshop of a ‘text engineer’
After some time, I am back with news from the world of Carolingian manuscripts. My critter this time is manuscript BPL 67F of the University Library of the University of Leiden, a collection of different glossographic texts from the late eighth or the early ninth century that came into being in northeastern France. Such collections were relatively common in the course of the eighth and the ninth centuries. They were, in fact, the most common form in which glossographic texts, i.e. texts concerned with words and their meaning, were transmitted in the Carolingian period. Perhaps, this form had to do with the ease of reference. Rather than to have multiple short, unbound libelli with word lists, or to compile these shorter texts into one large glossary, it was more practical to bind the short textual units together. In this manner it was possible to avoid loss (as short unbound booklets were prone to go missing), but also the time-consuming compilation involved in reshuffling the individual items in many short glossaries organized in different ways in order to feed one large compilation, say such that would involve advanced, i.e. at least partially alphabetical, order.
BPL 67F contains twelve such shorter glossographic texts in its 159 folia. Some of these, such as Affatim (fols. 1r-54r), Ab absens (fols. 54r-62r), Abavus (fols. 62r-104v), Arma (fols. 119r-128r), and the Glossae Nonii (fols. 142v-147r and 147r-148r) may be considered glossaries proper, that is they aim in different degrees to encompass the Latin language in totality and span across the whole of Latin vocabulary. Others, such as the Virgilian glosses (fols. 104v-118r), the Synonyms of Cicero (fols. 129r-139r, and 139v-141v), and the Glossae spiritales Eucherii (fols. 149r-152r) may be called rather glossographic texts, since they are concerned only with vocabulary found in a particular author (Virgil, Cicero) or a text (the Bible in case of Glossae Eucherii). The last type of material to be found in BPL 67F are the Voces variae animantium (fol. 152r), another type of a glossographic text, in this case one that covers only one register – the sounds made by various animals and objects. The very last item in BPL 67F, a text bearing title Glossae iurae (fols. 148v-149r), is not a glossographic text proper, although designated as such in the manuscript. The Glossae iurae are rather a list of juridical abbreviations used in legal manuscripts for abbreviations of technical vocabulary. In other manuscripts from the same period, lists such as this one were attributed to the grammarian Valerius Probus, or to Julius Caesar, and Isidore includes them into his Etymologiae as notae iuridicae.
The Synonyms of Cicero
In this piece, I wish to focus on one of the glossographic texts preserved in BPL 67F, the Synonyms of Cicero. The Synonyma Ciceronis can be described as a glossographic text in the form of clusters of words (i.e. batches) where all words in a batch represent synonyms or near synonyms taken (supposedly?) from the speeches of Cicero. For example, one such batch of synonymous words is orator. actor. defensor. patronus. causedicus. As the preface of this text claims, the Synonyma were compiled in order ‘to make the speech richer and its composition quicker by bringing together [similar] words’. In other words, the Synonyms were meant to be used as a sort of a proto-thesaurus.
The Synonyms of Cicero survive in multiple formats. Two versions are more wide-spread, while others are confined to a single manuscript, or in one case in two manuscripts. The two dominant versions are characterized by being ordered as glossae collectae (i.e., not in alphabetical order, in batches), beginning with the word orator, and in the A-order (i.e. according to the first letter of the word used as a head of the batch), starting with the word abditum.
The Synonyms are also equipped with a preface in some manuscripts, principally where the two dominant versions, orator and abditum, are present. This preface is disguised as a letter from Cicero to his friend Veturius in which Cicero instructs the latter how to use the text, claiming it is a useful tool for those ‘who strive to compose speeches’. Since there is no reason to believe that the Synonyms are in any way related to historical Cicero, this letter should be seen as a piece of legitimizing pseudepigraphy that throws light on the original function of the Synonyma. Based on the version present, the letter of ps-Cicero appears in two variants as well. In one, the last sentence of the letter reads: igitur ab oratore initium accipimus, indicating that the first batch that follows begins with the word orator, i.e. the Synonyms are organized as glossae collectae. In the other, the last sentence is: igitur per alfabetum initium capiamus, indicating that the batches in the Synonyms are organized in A-order and the following batch is abditum.
The Synonyms of Cicero in BPL 67F
BPL 67F holds pride of place among the manuscripts transmitting the Synonyms. It is one of the earliest witnesses of the letter of ps-Cicero. It is also one of three manuscripts written before 1000 that contain more than one version of this glossographic text. Interestingly, these two versions in BPL 67F are not identical. The first Synonyms (Synonyma I) are organized in batches, i.e. as glossae collectae; while the second Synonyms (Synonyma II) are organized in A-order. Both Synonyma I and Synonyma II are prefaced by the letter of ps-Cicero, and in both versions the same letter occurs, namely ab oratore which introduces normally the batched version beginning with orator. Indeed, orator is the first batch in both Synonyma I and II; however, while the next item in Synonyma I is dissertus. eloquens. facundus. ingeniosus, characteristic of the standard orator version, the Synonyma II continue with acer. intentus. vehemens. perseverans. indeclinatus, and then go on with other batches beginning with the letter A. Synonyma II are thus neither the standard batched version which begins with the batch orator and goes on in a particular order, nor is it the standard A-ordered version which begins with the batch abditum. They are rather a unique deviant A-ordered version which arose as a result of independent reshuffling of the batched orator version. This is clear from the presence of the ab oratore letter as the preface as well as from the presence of the item orator as the first batch in the Synonyma II.
Synonyma II stand out also upon comparison with other non-standard A-order versions. These do not include the letter of ps-Cicero, which was either never present in their batched prototype, or alternatively, which was present in this prototype (in ab oratore version) and was removed because it did not fit the reshuffled A-ordered text. Instead, the compiler of Synonyma II chose to retain the letter, which was present in the source of his compilation, and retained also the orator batch as the first item in the otherwise A-ordered Synonyms. This shows that his act of re-compilation was idiosyncratic and employed a unique method of re-shuffling the material.
Furthermore, this transformation from batched to A-ordered Synonyma affected not only the order of the items but also their layout on the page. Synonyma I are laid in four columns so that each word occupies a new line. In this way, the text was supposed to be consulted by columns rather than by lines. This is also the format of the standard orator version. On the contrary, although Synonyma II are also laid in four columns, the items in each batch are organized in lines in the manner of glossae, i.e. with one of the words as a lemma, and the other organized following it as an interpretamentum. The reading direction of this version was horizontal rather than vertical. The line-by-line format can be encountered also in other A-ordered exemplars of the Synonyma.
The transition from batches to A-order was not only matter of organization, but it seems, also had to do with the intended use and signification. Arrangement in columns is typical for differentiae, a genre of grammatical texts which listed ‘false friends’ or near-synonyms, and which belonged into the realm of the grammatica. However, arrangement in lines is a feature of glossaries and glossographic collections. While Synonyma I then continued to resemble differentiae, Synonyma II were ‘upgraded’ to look like a glossographic text.
The Synonyms of Cicero came into being as differentiae, perhaps a classroom text intended to aid one in the composition of a speech. However, this is not the context, nor the purpose of the Synonyms preserved in BPL 67F. In this manuscript, they were transformed by their insertion into a glossographic collection into a piece of glossography. In the case of Synonyma II, such re-framing went even further. It did not involve only re-contextualization, but also re-organization of the content from a non-alphabetical order to the A-order, and a change in layout towards what was the standard format of a glossographic text. Synonyma II in BPL 67F are one of the earliest exemplars of the Synonyms of Cicero which underwent these transformations, the early swallow that heralded the trend which would set in in the course of the ninth century.
In addition, Synonyma II display peculiar eccentricities, such as the presence of the ab oratore letter preface in the A-ordered text, which do not have parallel elsewhere in the textual history of the Synonyms of Cicero. These peculiarities indicate that the transformations were idiosyncratic. The second version of the Synonyms of Cicero in BPL 67F reveals, in other words, working of an independent agent who made his own choices in order to fit the text into a new, glossographic context. BPL 67F allows us to observe this medieval compiler at work in his textual workshop, since he left many traces of his activity on the text.
BPL 67F is certainly not the only glossographic collection which was a product of the medieval ‘textual engineering’. Quite the contrary, we should imagine that the majority of the early medieval collections akin to BPL 67F came into being in the same fashion, ‘engineered’ by anonymous compilers who endowed old texts with new meanings and significance which suited their transforming intellectual horizons. The uniqueness of BPL 67F rests in the fact that it is one of the oldest such collections and that the traces of re-working are more visible in this manuscript than in other glossographic collections from the same period.
The transition from one frame of significance to another that I illustrated by the Synonyms of Cicero as preserved in BPL 67F did not happen overnight. We can pinpoint a moment in time when a particular text (the Synonyms of Cicero) was perceived overwhelmingly as one thing (i.e. a set of differentiae), and another moment when it was perceived primarily as something different (i.e. as a glossographic text). The transition from one stage of existence to another involved many small steps, each of which pushed the Synonyms further away from the realm of grammatica and closer to the realm of glossography. It is impossible even to identify the tipping point when the Synonyms of Cicero were perceived more often as one thing rather than as something else. In the textual history of this text, and others similar, there is never a stage about which we can say that the Synonyms functioned only as a grammatical aid, or only as a glossographic text. Rather, we should think in terms of ratios, contexts, and frameworks of significance. The Synonyma Ciceronis as preserved in BPL 67F, then, reflect first and foremost a particular trend, a trend which gave rise to a whole early medieval genre of technical literature – the glossographic collection.
 Bernhard Bischoff, Katalog der festländischen Handschriften des neunten Jahrhunderts (mit Ausnahme der wisigotischen). II: Laon-Paderborn, Bayerische Akademie der Wissenschaften. Veröffentlichungen der Kommission für die Herausgabe der mittelalterlichen Bibliothekskataloge Deutschlands und der Schweitz (Wiesbaden: Harrassowitz, 2004), n. 2142. Digital images of selected pages are available at: https://socrates.leidenuniv.nl/R/-?func=dbin-jump-full&object_id=673839.
 Rosamond McKitterick, ‘Glossaries and Other Innovations in Carolingian Book Production’, in Turning over a New Leaf: Change and Development in the Medieval Book, ed. Rosamond McKitterick, Erik Kwakkel, and Rodney Thomson (Leiden, 2012), 70. McKitterick calls such collections ‘glossary chrestomathies’.
 Since a majority of glossaries and glossographic texts does not have a title, they are generally named after their first lemma (and thus their names most commonly begin with the A). This will be also my standard manner of reference to them, including in the cases when multiple versions of one text exist which can be distinguished on the basis of the sequence of the words they contain.
 They can be found edited in Georg Goetz, Corpus glossariorum latinorum, 7 vols. (Leipzig: Teubner, 1888)., volumes 4 (Abavus, Ab absens, Affatim) and 5 (Glossae Nonii). The glossary Arma was not edited by Goetz, but is discussed in detail, together with some of the other glossaries mentioned here, in Anna Carlotta Dionisotti, ‘On the Nature and Transmission of Latin Glossaries’, in Les Manuscrits de Lexiques et Glossaires de L’antiquité Au Moyen Âge,, ed. J. Hamesse (Louvain, 1996), 202–252. The glossaries of BPL 67F are also discussed in Georg Goetz, De glossariorum latinorum origine et fatis (Leipzig: Teubner, 1928), 129–141.
 The Glossae Virgilianae are edited in Corpus Glossariorum Latinorum, volume 4.
 Only some versions of the Synonyma Ciceronis have been critically edited so far, probably because they survive in multiple text versions and relationships between these versions are complex and unclear. The two principal versions were re-printed from fifteenth- and sixteenth-century early prints in the nineteenth century as Wilhelm Leonhard Mahne, ed., M. Tullii Ciceronis quae vulgo feruntur Synonyma ad Lucium Veturium secundum editiones Romanas denuo excudi (Brill, 1850). and Wilhelm Leonhard Mahne, ed., M. Tullii Ciceronis (quae Vulgo Feruntur) Synonyma Ad Lucium Veturium Secundum Editionem Parisinam Denuo Excudi (Leiden: Brill, 1851). Paolo Gatti has edited two other, isolated versions as Paolo Gatti, Synonyma Ciceronis (arba, humus) (Universita di Genova, Facolta di lettere, 1993). and Paolo Gatti, Synonyma Ciceronis. La Raccolta ‘Accusat, lacescit’ (Trento: Dipartimento di scienze filologiche e storiche, 1994). Barwick edited another isolated version in his edition of the Bobbio Charisius; Flavius Sosipater Charisius, Artis Grammaticae Libri v, ed. Charles Barwick, Bibliotheca Scriptorum Graecorum et Romanorum Teubneriana 1137 (Leipzig: Teubner, 1964), 412–449.
 This glossographic text was recently edited as Lugdunensis Eucherius, Eucherii Lugdunensis Formulae spiritalis intellegentiae, ed. C. Mandolfo, Corpus Christianorum Series Latina 66 (Turnhout: Brepols, 2004).
 No critical edition of the Latin text was made so far. A study on the voces literature in different languages, however, was published as Wilhelm Wackernagel, Voces Variae animantium: ein Beitrag zur Naturkunde und zur Geschichte der Sprache (Basel: Bahnmaier (Detloff), 1869).
 Various versions of this list of juridical abbreviations were edited in Heinrich Keil, Grammatici Latini, vol. 4 (Hildesheim: Olms, 1961), 265–352.. This particular version of the notae iuris are transcribed in pp. 277-81.
 Etym. 1.23.1: Quaedam autem litterae in libris iuris uerborum suorum notae sunt, quo scriptio celeris breuior que fiat; according to Wallace Martin Lindsay, Etymologiarum Sive Originum Libri XX, vol. 1, 2 vols. (Oxford: Clarendon Press, 1911).
 The Synonyms of Cicero are discussed in Goetz, De glossariorum latinorum origine et fatis, 75–86. and in Giorgio Brugnoli, Studi sulle Differentiae verborum (Rome: Agnelo Signorelli, 1955), 27–37. Gatti believes that the different versions of the Synonyms are in fact unrelated and the attribution to Cicero is a generic phenomenon, Gatti, Synonyma Ciceronis. La Raccolta ‘Accusat, lacescit’, 8. As with many texts, the origin of the Synonyma Ciceronis is unknown and the earliest manuscript evidence we have for them comes from the eighth century.
 Collegi haec verba… quo uberior promptiorque fiat oratio verborum consertione; see Brugnoli, Studi sulle Differentiae verborum, 28.
 The most complete overview of manuscripts containing the Synonyms of Cicero can be found in Birger Munk Olsen, L’étude des auteurs classiques latins aux XIe et XIIe siècles, vol. 1, Documents, études et répertoires publiés par l’Institut de Recherche et d’Histoire des Textes (Paris: CNRS, 1982), 133–134 and 340–350. Brugnoli also lists some of the manuscripts, but his list is not as complete as Munk Olsen’s; see Brugnoli, Studi sulle Differentiae verborum, 27–37. All in all, Munk Olsen records twenty-eight manuscripts containing one or in some cases two exemplars of the Synonyms from before 1000. Five of these are dated to the eighth or perhaps early ninth century and other twenty-two are from the ninth and tenth century.
 The two dominant versions, orator and abditum appear in all but seven of the manuscripts listed by Munk Olsen. The six ‘deviant’ versions surviving in these seven manuscripts are: Amor.ardor preserved in Naples, Biblioteca Nazionale, MS IV.A.8 (8th century, Bobbio); Arba. hummus in London, British Library, MS Harley 5792 (8th/9th century, Italy); Acer. intentus in BPL 67F discussed here (8th/9th century, northeastern France); Inanis.longus surviving in two perhaps related manuscripts, Brussels, Koninklijke Bibliotheek, MS 9311-19 (9th century, 1/3, area of St. Amand) and Bern, Burgerbibliothek, MS 224 (9th century, 1/3, France); Accusat. lacessit in Leiden, Universiteitsbibliotheek, MS BPL 67E (9th century, in., Lotharingia/Upper Rheinland); and Auctor.orator in Montpellier, Bibliothèque Universitaire. Section Médecine, MS H 160 (9th century).
 This letter preface was edited by Brugnoli; Brugnoli, Studi sulle Differentiae verborum, 28–29 and 31.
 The context one can imagine is that of rhetoric, perhaps classroom composition. Brugnoli, however, expressed opinion that the preface might be a younger addition to the Synonyma as it features only in some manuscripts of the Synonyms, and also because it seems to be tied specifically to the two major versions; Brugnoli, Studi sulle Differentiae verborum, 29.
 The oldest witness of the letter is Albi, Bibliothèque Municipale, MS 29 (115) from the second half of the eighth century, perhaps not significantly older than BPL 67F. The Albi manuscript is digitized at: http://mediatheques.grand-albigeois.fr/1035-manuscrits-medievaux.htm.
 The two other manuscripts are London, Harley 5792, digitized at: http://www.bl.uk/manuscripts/FullDisplay.aspx?ref=Harley_MS_5792; and Paris, Bibliothèque Nationale de France, MS Lat. 2341 (9th century, 2/4, France), digitized at: http://gallica.bnf.fr/ark:/12148/btv1b85409594. In the latter manuscript, the two versions present are both prefaced by ab oratore letter and are the standard batched version (orator); see Olsen, L’étude des auteurs classiques latins aux XIe et XIIe siècles, 1:347. The British library manuscript does not contain the preface; see Ibid., 1:340.
 Brugnoli reports that the manuscript Vat. Lat. 3321 also contains a ‘deviant’ version of the Synonyms beginning with acer. intentus, but I was not able to verify it against the volume 4 of Corpus glossariorum latinorum to which he refers; see Brugnoli, Studi sulle Differentiae verborum, 33. Elsewhere, Brugnoli mentions that the manuscript contains only a single item from the Synonyms added in a second hand.
 Note the peculiarity in the oldest witness of the Synonyms of Cicero, the Bobbio Charisius. the Synonyms are introduced in the contents page of this grammatical compendium as Synonyma Ciceronis indifferenter, but the incipit of the text reads Synonyma Ciceronis ordine litterarum conposita; Charisius, Artis Grammaticae Libri v, 412.
 See for example Munich, Bayerische Staatsbibliothek, MS Clm 14388 (mid-9th century, Northwestern Germany); at: http://daten.digitale-sammlungen.de/bsb00022465/image_465; Brussels 9311-19 (9th century, 1/3, area of St. Amand), at: http://lucia.kbr.be/multi/KBR_9311-19Viewer/imageViewer.html (from fol. 128v); or Paris Lat. 2341, at: http://gallica.bnf.fr/ark:/12148/btv1b85409594/f544.zoom.r=2341.langFR.
 The Synonyma II in BPL 67F are the oldest exemplar of the Synonyms of Cicero organized in this manner. Other examples include Leiden, Universiteitsbibliotheek, MS Voss. Lat. F 24 (9th century, 2/4, area of Tours); Leiden, Universiteitsbibliotheek, MS BPL 67E (9th century, in., Lotharingia/Upper Rheinland); and Munich, Bayerische Staatsbibliothek, MS 14252 (9th century, ¼; St. Emmeram, Regensburg), at: http://daten.digitale-sammlungen.de/bsb00046836/image_355.
 This ‘genre’ is described and discussed in Goetz, De glossariorum latinorum origine et fatis, 87–93; and in Brugnoli, Studi sulle Differentiae verborum.
 Note that BPL 67F contains one item which was non-glossographic, even non-lexical, in character – the list of legal abbreviations. The presence of the notae iuris in a glossographic collection like BPL 67F, and its rebranding as glossae, further supports the impression that BPL 67F, and other collections similar to it, absorbed material from diverse areas and transformed it progressively into glossographic material.
 Even after their intrusion into the realm of glossography, the Synonyms of Cicero continued to be transmitted as a grammatical text in grammatical compendia. This is the context of their transmission in Paris, Bibliothèque Nationale de France, MS Lat. 7520 (10th/11th c.?, France?), at: http://gallica.bnf.fr/ark:/12148/btv1b84900632/f153.item; and in Oxford, Bodleian Library, MS Add. C. 144 (11th c., in., Central Italy); see Martin Irvine, The Making of Textual Culture: ‘Grammatica’ and Literary Theory 350-1100, Cambridge Studies in Medieval Literature 19 (Cambridge: Cambridge University Press, 1994), 353. The oldest surviving manuscript containing the Synonyma Ciceronis is also a grammatical compendium, not a glossographic collection: Naples, Biblioteca Nazionale, MS IV.A.8 (8th century, Bobbio).
Citation: Evina Steinova: Carolingian Critters IV: Leiden, Universiteitsbibliotheek, BPL 67F. A peep into the workshop of a ‘text engineer’, in: Mittelalter. Interdisziplinäre Forschung und Rezeptionsgeschichte, 30 January 2014, http://mittelalter.hypotheses.org/2929.
[…] Usually it concerns a dictionary or encyclopedia (read about an example of four columns here). A page with more than four columns is extremely rare. In fact, I only know of one such manuscript […]