The Codex Pagesianus (BAV, Pagès 1) and the emergence of Aristiotle in the medieval west 8821010724, 9788821010729

Citation preview

THE CODEX PAGESIANUS (BAV, PAGÈS 1) AND THE EMERGENCE OF ARISTOTLE IN THE MEDIEVAL WEST

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 1

27-Sep-21 8:05:50 PM

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 2

27-Sep-21 8:05:50 PM

STUDI E TESTI ———————————— 549 ————————————

JOHN MAGEE – FABIO TRONCARELLI

THE CODEX PAGESIANUS (BAV, PAGÈS 1) AND THE EMERGENCE OF ARISTOTLE IN THE MEDIEVAL WEST

C I T T À D E L VAT I C A N O B i B l i o t e c a a p o s t o l i c a V at i c a n a 2022

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. 00 Frontmatter.indd 3

29-Nov-21 10:54:41 PM

La collana “Studi e testi” è curata dalla Commissione per l’editoria della Biblioteca Apostolica Vaticana

—————— Proprietà letteraria riservata © Biblioteca Apostolica Vaticana, 2022 ISBN 978-88-210-1072-9 Edizione digitale: ISBN 978-88-210-1073-6 www.vaticanlibrary.va/it/pubblicazioni

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 4

27-Sep-21 8:05:50 PM

TABLE OF CONTENTS Foreword The Codex Pagesianus (BAV, Pagès 1) and the Emergence of Aristotle in the Medieval West.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Chapter I The Codex Pagesianus and the Early Textual Tradition of Boethius’ First De interpretatione Commentary . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Chapter II The Extant Tradition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 Chapter III Edited Passages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 Chapter IV Arbeo, Leidrat, Alcuin and the Codex Pagesianus . . . . . . . . . . . . . . . . . 49 Chapter V The Devil in the Mirror: Alcuinian Drawings in the Codex Page­sia­nus . 131 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 5

27-Sep-21 8:05:50 PM

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 6

27-Sep-21 8:05:50 PM

FOREWORD

THE CODEX PAGESIANUS (BAV, PAGÈS 1) AND THE EMERGENCE OF ARISTOTLE IN THE MEDIEVAL WEST The present volume seeks to shed light on developments that led to the rediscovery of Aristotle in the medieval West. More precisely, it is an investigation of the oldest extant manuscripts of Boethius’ elementary commentary on De interpretatione (c), which embodied a partial Latin translation of Aristotle’s treatise and was accompanied in some manuscripts by Boethius’ translations of the Categories and (or) Isagoge. The termini a quo and ad quem perhaps most relevant to this study are 524/5, the date of Boethius’ death, after which Aristotle gradually fades from view in the Latin West, and 784-98, the period during which the Codex Pagesianus (BAV, Pagès 1) was produced. The Pagesianus sends the first known signals of scholarly activity that put an end to two centuries of philosophical silence in the West, and with it and its younger contemporaries the “Old Logic” first becomes visible to us today. The emergence of this logical corpus reflects both dynamic and static modes of scholarly engagement with the philosophical texts contained therein, in that our oldest manuscripts bear traces of evolving as well as fixed and older strata of editorial intervention. In this sense, Aristotle did not appear ex nihilo, instantly and fully formed, in the early medieval West, although the story recounting the movement of his Categories and De interpretatione from the second half of the 6th c. to the third quarter of the 8th remains obscure and for the most part undocumented. Boethius’ father-in-law Symmachus (d. 525/6), a minor aristocrat named Martius Novatus Renatus, a scribe and student of Priscian named Theodorus (fl. later 520s-?), the politician and monk Cassiodorus (d. c. 585) — these and unnamed others variously contributed to the preservation of Boethius’ works, evidence for whose survival in the 7th and earlier 8th cc. is however indirect, with the exception of a lone fragment of De arithmetica, and relatively negligible. The corpus appears to have moved whole or in part from Ravenna to Constantinople and then to Calabria during the second half of the 6th c., but to trace its subsequent migration from these and other regions to Lombardy and intellectual centres north of the Alps in the later 8th c. is a challenging task, to say the least.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 7

27-Sep-21 8:05:50 PM

8

Foreword

Despite the uncertainties, we may confidently state that the transmitted text of c descends from an archetype (ω) that carried numerous errors and had certain distinctive paratextual features. In what follows, it will be provisionally traced back to Vivarium, c. 560-80. The path from ω leads first, among extant witnesses, to the Pagesianus, whose text, like those of its younger contemporaries, reflects significant contamination at least some of which must predate 784, or the Pagesianus itself. Indeed, comparison of the Pagesianus (V) with its closest congeners (CK) reveals two lost intermediaries, the parents of CV (δ) and of Kδ (μ), both situated in a branch of the textual tradition (ς) in which gaps between lemmata were left untouched — in accordance with Boethius’ own intentions — by passages excerpted from his continuous translation of De interpretatione (t); a third intermediary (ζ) in this non-interpolated branch must also be postulated to explain certain symptoms loosely shared by three other manuscripts (DFP). The interpolated branch of the tradition (φ), on the other hand, is defined not only by passages excerpted from t to produce lemmata covering the whole of De interpretatione but by a host of errors and variants reflecting additional processes of collation, correction, and other scholarly activity. Lorenzo Minio-Paluello observed long ago that the text of t suffered contamination early on, tracing its first intermingling with the “Alcuinian corpus” of which c formed a part to Corbie, c. 820.1 As we hope to show, however, the Pagesianus contains evidence indicative of older strata of contamination that date back to μ and may well have originated significantly south of Corbie. Echoes issuing from the ς-witnesses suggest, moreover, that some of the scholarly activity ultimately responsible for the contamination reaches even further back, to ς. Hence the Pagesianus, despite its age and authority, cannot be deemed a codex optimus but must be weighed against other witnesses on both sides of the tradition if we are to gain clarity on the progress of Boethius’ ipsissima verba through time and place. There is perhaps an element of irony in the fact that the Pagesianus, our first direct sighting of the medieval Aristotle, contains so little of Aristotle himself, but such is evidently the nature of some momentous first-beginnings. Although the Pagesianus has been claimed by scholars of considerable authority to descend from an “archetype” or exemplar possessed by Alcuin, there are reasons for regarding this view as problematical, as will become clear in the pages that follow. To begin with, the composition of the Page­ sianus should in our opinion be dated prior to 798, the year of Leidrat’s ordination as bishop of Lyon, and that on the basis of its scripts, which do not comport with graphic experimentation known to have taken place at 1 Minio-Paluello

[ed.], Arist. Lat. II, p. xlii.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 8

27-Sep-21 8:05:51 PM

Foreword

9

the end of the 8th c. and turn of the 9th but which bear close resemblance to certain eclectic and disorderly scripts datable to the second half of the 8th c. In addition, its flyleaf contains a hitherto undetected note of possession, evidently an autograph, written in a hand characteristic of the end of the 8th c., indicating that at least the first booklet of the Pagesianus once belonged to Leidrat’s teacher, bishop Arbeo of Freising (d. 784), who is also recalled by two further notes of ownership written in contemporary hands on the front and back of the same leaf. For the first booklet of the Page­ sianus, then, both script and notes of possession point to Freising and the episcopate of Arbeo. There is in addition a dedicatory inscription, known already to Léopold Delisle, which establishes a connection with Lyon and the episcopate of Leidrat roughly a decade and a half later; it appears on f. 1v and is conveniently accessible now in CLA IV (417). Since each of the works contained in the Pagesianus has a fascicle numbering that was later corrected in order to produce a continuous series, it is natural to interpret Leidrat’s notes of possession as successive ex libris independently added to individual booklets. Later booklets were written by the same scribes as wrote the first booklet but in different inks, as occurs when texts are transcribed at intervals rather than in immediate succession. Certain pages, moreover, were written by other copyists in a script resembling that employed by the first group; this handwriting is reminiscent of one also found in certain Bavarian manuscripts of the same era. From the combined evidence of the notes of possession we may surmise that the Pagesianus was written intermittently over an extended period of time by copyists who faithfully followed Leidrat through the course of his career from Freising to Lyon. In this sense, the Pagesianus may be deemed a “fictitious” codex, an artificial collection comprised of disparate parts and transcribed in stages by a single group of copyists between 784 and 798. It combines what at the time must have appeared unusual and exotic works of logic, works that had originally been copied separately but were subsequently assembled into a single volume in consideration of the deep connective tissue binding them. Thanks to Hilduin, Eriugena, and Anastasius Bibliothecarius the Carolingians succeeding in bringing the Corpus Dionysiacum to the Latin West; for their knowledge of Aristotle, by contrast, they remained confined to Boethius’ translations and commentaries, and it was not until the reign of Otto III (983-1002) that systematic engagement with the Boethian translations gained significant momentum, as is vividly illustrated by the versions of the Categories and De interpretatione produced in St. Gall by Notker the German, who blended his own Alemmanic dialect with Boethius’ Latin. To judge from the evidence presented for consideration in the present

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 9

27-Sep-21 8:05:51 PM

10

Foreword

volume, it appears that the Carolingian rediscovery of Aristotle — more precisely, the propagation of what later became known as the “Old Logic” — ultimately depended on Alcuin’s filius Leidrat rather than on Leidrat’s spiritual pater Alcuin. The rediscovery proper appears to have occurred in Leidrat’s home territory of Noricum, which embraced both Arbeo’s Freising and the Salzburg of bishop Virgilius, who may have had no faith in the Antipodes but who nevertheless studied rare works with a dedication sufficient to earn him the nickname “the Geometer.” It also involved parts of Lombardy, with its libraries in which ancient authors such as Boethius had been cultivated for ages, as at Bobbio, once the home of the Turin De arithmetica fragment and the place where Gerbert of Aurillac encountered a copy of Boethius’ De astronomia, a work irrevocably lost to us today.2

2 The

authors bear joint responsibility for the Foreword and for development of the volume’s theme. Otherwise, JM is responsible for cc. I-III, and FT for cc. IV-V. The authors would like to thank Anthony Fredette, Roberto Granieri, and Francesco Pica for their capable editorial support, and Dr. Paolo Vian, Dr. Marco Buonocore, and other members of the BAV staff for their encouragement and support.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 10

27-Sep-21 8:05:51 PM

CHAPTER I

THE CODEX PAGESIANUS AND THE EARLY TEXTUAL TRADITION OF BOETHIUS’ FIRST DE INTERPRETATIONE COMMENTARY Boethius’ death in 524/5 brought a violent and untimely end to his project, formally articulated approximately a decade earlier, of translating, commenting on, and harmonizing all of the works of Plato and Aristotle.1 To judge only from what has survived, in the period roughly between his consulship and imprisonment (510-23) he completed translations of the Isagoge, Categories, De interpretatione, Prior Analytics, Topics, and Sophis­ tical Fallacies, and commentaries on the Isagoge, Categories, and De inter­ pretatione.2 Of the philosophical translators who preceded him, Marius Victorinus and Vettius Agorius Praetextatus, both active in the 4th c., obviously stand out, although their accomplishments are virtually unknown, for of Victorinus’ translations there survive only the passages of the Isagoge preserved by Boethius in his first commentary on the same, while Praetextatus translated Themistius’ paraphrases of the two Analytics and may have made or motivated the Latin paraphrase of the Categories — apparently of Themistian origin as well — which Alcuin later misattributed to Augustine.3 After Boethius, the project of translating Aristotle into Latin 1 The following study develops preliminary findings articulated in Magee, Observations (corrections to typographical errors, p. 15 [In Perih. I 49.20, 50.3]: separatumque, compo­ siti ­ sque). Boethius’ De interpretatione commentaries are cited according to the editions of C. Meiser (Leipzig 1877, 1880; remarks on the first at Magee, Observations, p. 13); his continuous translation is cited according to the traditional Bekker numbers or (and) L. Minio-Paluello’s edition in Arist. Lat., II.1 (Bruges – Paris 1965); Aristotle’s De inter­ pretatione is cited as from H. Weidemann’s magisterial edition (Berlin – Boston 2014). JM wishes to express his gratitude to Dr. Paolo Vian (Biblioteca Apostolica Vaticana) and Dr. Timoty Leonardi (Biblioteca e Archivio Capitolare, Vercelli), to Prof. Gyburg Uhlmann and Dr. Christian Vogel (Freie Universität, Berlin), to Profs. George Boys-Stones and Phillip Horky (Durham University), for their generous support and responses to material developed in preparation for this study, and to Anthony Fredette, Roberto Granieri, and Francesco Pico (University of Toronto) for their editorial support. 2 Cfr. Magee, Composition, pp. 10-23; Boethius, pp. 790-793; Magee-Marenbon, App­en­ dix, pp. 304-305; Gruber, Boethius, pp. 26-44. 3 Arist. Lat., I.1-7 (cfr. Hadot, Victorinus, pp. 367-380); Boeth., In Perih. II 3.7-4.3. The Ω-redaction of Cassiodorus’ Institutiones misattributes Boethius’ translations of Aristotle

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 11

27-Sep-21 8:05:51 PM

12

chapter i

languished until James of Venice’s Posterior Analytics (1128), which ushered in the golden age of John Sarracen, Robert Grosseteste, and William of Moerbeke. Hence the Carolingians, who through the agency of Hilduin, Eriugena, and Anastasius Bibliothecarius introduced the Corpus Dio­ nysiacum to the Latin West, were confined to Boethius’ translations and commentaries for their knowledge of Aristotle, and it was not until the reign of Otto III (983-1002) that systematic engagement with the Boethian translations gained significant momentum, as is perhaps illustrated most concretely by the macaronic versions of the Categories and De interpreta­ tione that Notker the German compiled in St. Gall, freely interlacing his own Alemmanic dialect with Boethius’ Latin. The present investigation builds on preliminary conclusions reached in an earlier study4 and on the analysis of the Codex Pagesianus that follows, and seeks to shed light on developments that led to Aristotle’s first appearance in the medieval West — more precisely, to his first moment of visibility, direct or indirect, in the manuscripts that have survived. It consists of three parts (cc. I-III). The first treats of the fate of the Boethian corpus, c. 550-785, from which almost no manuscript evidence survives (c. I). The second part focuses on eleven of the earliest extant manuscripts of his elementary De interpretatione commentary, including certain aspects of their mise-en-page that evidently harken back to the oldest stages of transmission (c. II). The manuscripts are5: C D F G H K L M O P V

Paris, BnF, lat. 13956, 9th c. Paris, BnF, lat. 6288, 10th-11th c. Munich, Bayer. Staatsbibl., clm 6374, 9th c. St. Gall, Stiftsbibl., 820, 9th-10th c. Paris, BnF, lat. 6400F, 11th-12th c. Kraków, Bibl. Jagiell., Berol. Lat. Qu. 687, 9th c. Philadelphia, Univ. of Penn., Schoenb. 101, 9th/11th cc. Munich, Bayer. Staatsbibl., clm 14377, 10th-11th c. Oxford, Bodl., Laud. lat. 49, 11th c. Paris, BnF, lat. 12960, 9th c. Vatican City, BAV, Pagès 1, c. 785

and Porphyry to Victorinus (2.3.18); cfr. Courcelle, Brouillon, pp. 82-84 (94-96); Pecere, Cassiodoro, pp. 193-194. 4 Magee, Observations. 5 K has been newly introduced, and the following four sigla changed from those listed parenthetically: C (Pa); L (Lo); O (Ox); V (R). L is a composite and highly confused witness: (a) fols. 2-4 (In Perih. I 31.1-36.17, Magna quidem … species sunt) date to the 11th c.; (b) fols. 5-44 (In Perih. I 36.17-111.20, rectum est … non contingent), to the 9th c.; (c) fols. 45-53 (In Perih. I 160.15-225.14, praedicari non … series explicabit), to the 11th c. Its changes of hand, block dislocations of text, and lacunae render the discernment of any pedigree difficult if not impossible.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 12

27-Sep-21 8:05:51 PM

the early textual tradition

13

Finally, the third part consists of two passages newly edited from the manuscripts and supplemented by brief textual commentary (c. III). The Lost Tradition The path traveled by Boethius’ writings and library during the second quarter of the 6th c. raises a number of questions of considerable importance. Manuscript subscriptions indicate that Boethius revised his own copies of De arithmetica and the commentary on Cicero’s Topics, while other works make it clear that he also solicited revisions from Symmachus, John the Deacon, and Patricius.6 De arithmetica is of course an early work (c. 500-06), but the Cicero commentary (c. 520-23) indicates that Boethius continued to edit and revise despite the mounting pressures7 of the years leading into his period as Master of the Offices (522-23). Although the arrest and incarceration in Verona and then Pavia obviously put an end to Boethius’ access to his library, the continued availability of some books cannot be ruled out a priori.8 Symmachus evidently revised the text of the Consolatio and may have cared for the library until his own death in 525/6, after which Rusticiana and other relations or friends such as Martius Novatus Renatus (on whom more below) presumably took over.9 In 527 Amalsuintha restored the confiscated estates of the Anicii in a desperate effort to foster goodwill between Romans and Ostrogoths,10 but any contemporary records that may have documented the events of 523-27 were so effectively buried that it is impossible to say precisely what became of Boethius’ library and writings during those years.11 Similarly, it is uncertain whether any of his works or portions of his library ever reached the university library established in Rome by Cassiodorus and Agapetus I in 535-36 or the Lateran library thereafter.12 The first unmistakable signs of survival date to the second half of the 6th c. Two of them are relatively uninformative, consisting of faint echoes of the Consolatio in Maximian’s Elegies and an Uncial fragment of De 6 Magee,

Text, p. 7, nt. 18 (the subscription, conditor operis emendavi, survives also in Valenciennes, Bibl. Mun. 406, ff. 44v, 49v); [ed.] De divisione, p. xxi, nt. 19; Vitiello, Tracce, p. 241; Theodahad, p. 86. 7 Cfr. Boeth., In Cat. 201b (with Cic., Divin. 2.1.1). 8 Magee, Gorgias, pp. 27-28; Troncarelli, Fiamma, p. 6 (and pp. 30-31). 9 Chadwick, Boethius, pp. 254-257; Troncarelli, Fiamma, pp. 163-167. 10 Vitiello, Amalsuintha, p. 86. 11 Cfr. Vitiello, Accusarentur, p. 352 ; Troncarelli, Afterword, pp. 528-533. 12 Cfr. Courcelle, Writers, pp. 334-335, 361-409; Troncarelli, Vivarium, pp. 7-15; Giuliani – Pavolini, ‘Biblioteca’.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 13

27-Sep-21 8:05:51 PM

14

chapter i

­ari­thmetica now in Turin.13 Far more significant is the evidence provided by Cassiodorus’ Institutiones, the second book of which in its various redactions pinpoints to the Vivarium library copies of Boethius’ translations of Nicomachus, Euclid, Aristotle, and Porphyry, his commentaries on Aristotle, Porphyry, and Cicero, and his De hypotheticis syllogismis and De topicis differentiis.14 From the evidence of the Institutiones in particular it is possible to trace the movement of Boethius’ works to Constantinople, during the period of Cassiodorus’ exile (c. 540-54), and then to Calabria, with his return. Given the unlikelihood that the corpus remained either fully intact or preserved in a unique copy during the decades immediately after Boethius’ death, Constantinople cannot be assumed to have been its sole haven during the Gothic War (535-55),15 and there can be no certainty about the precise date or means of its conveyance to the East, but that Cassiodorus and Vivarium played a crucial role in the early history of its transmission in the West is beyond doubt. Hermann Usener was the first to note (1877) that the second book of the Institutiones is transmitted in more than one recension. Research subsequently conducted by R.A.B. Mynors, Pierre Courcelle, and others revealed that passages of Boethius’ De topicis differentiis were interpolated by Cassiodorus or (and) his disciples in what have come to be known as the Φ- and Δ-redactions of the Institutiones.16 Also interpolated in the Δ-recension was a fragment whose initial discovery by Alfred Holder in fact occasioned Usener’s observation concerning the textual tradition. That fragment is the Ordo generis Cassiodororum or so-called Anecdoton Holderi, a genealogical review, excerpted from a letter addressed by Cassiodorus (c. 550-53) to Rufius Petronius Nicomachus Cethegus (cos. 504), which in certain details echoes a Vita Boethii that survives in some manuscripts of the Consolatio. The relationships between these different texts may be described as follows17:

13 Gruber, Boethius, p. 98; Vitiello, Theodahad, pp. 83-84; Codd. Boeth., III.338 (CLA, IV.450); cfr. Troncarelli, Fiamma, pp. 1-6, on a possible remnant of Boethius’ library (Verona Euclid, CLA, IV.501). A papyrus copy of the Cicero commentary is reported by Lupus of Ferrières (Ep. 53) to have been at Tours in the 9th c.; like the fragment of a papyrus codex of Avitus’ homilies now preserved in Paris (CLA, V.573), it may have dated to the 6th c. 14 Cassiod., Instit. 2.3.18, 4.7, 6.3. 15 Cfr. Troncarelli, Afterword, pp. 538, 540. 16 Courcelle, Brouillon, pp. 68-69 (80-81); Pecere, Cassiodoro, p. 195; cf. Troncarelli, Vivarium, pp. 11-21. 17 For the stemma of the Institutiones, see Courcelle, Brouillon, p. 85 (97); cfr. Galon­ nier, Anecdoton, pp. 10-14; Pecere, Cassiodoro, pp. 188-192. Editions: Instit. [ed. R.A.B. Mynors, Oxford 1937; cfr. Pecere, Cassiodoro, p. 195]; Top. diff. [ed. D.Z. Nikitas, Paris-

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 14

27-Sep-21 8:05:51 PM

15

the early textual tradition

Cassiod., Instit. (2.3.15-16)

Boeth., Top. diff. (1.2.1-7.28, 2.2.7-12.3, 4.5.6-12.12)

ω Ω

[Cassiod.]

Ordo generis

Vita Boethii

φ Φ Δ

The biographical notices for Symmachus and Boethius are chronologically ordered (Symmachus – Boethius) in the Ordo generis, as suits a genealogical review, but reversed (Boethius – Symmachus) in the Vita Boethii, as would have suited the prolegomena to an ancient edition of the Consolatio, one which Fabio Troncarelli has traced back to Cassiodorus (Constantinople, c. 538-54).18 Otherwise, they echo one another almost verbatim.19 Material that is closely related, and at points identical, to De topicis ­differentiis similarly found its way under separate titles into an ancient copy of Boethius’ logico-rhetorical monographs, the index to which has been variously preserved by a number of the oldest surviving manuscripts20: Quae sint in hoc codice Anicii Manlii Severini Boethii v. c. et ill. excons. ord.: I. De differentiis topicis libri IIII II. De divisione III. Communis speculatio de rhetoricae cognatione IIII. Locorum rhetoricorum distinctio V. De multifaria praedicatione VI. Quomodo argumentorum vel unde colliguntur loci VII. Liber ante praedicamenta VIII. Introductio in categoricos syllogismos, libri II VIIII. De hypotheticis syllogismis libri III Brussels 1990]; Ordo generis [ed. Galonnier (Ordo generis, pp. 306-307)]; Vita Boethii [ed. Troncarelli (Tradizioni, pp. 12-14)]. 18 Troncarelli, Tradizioni, pp. 1-106; Cogitatio, pp. 27-96; Umanesimo, pp. 88-97; Fiam­ ma, p. 43. 19 The Boethius notice is omitted by three manuscripts of the Vita Boethii (12.1-7 T. = Ordo generis 9-14), and the Symmachus notice by two (13.13-17 T. = Ordo generis 5-8); the reversal in order and these block omissions are telling indications of the antiquity of the notices (cfr. below, nt. 39). 20 Magee, Text, pp. 4-7; [ed.] De divisione, p. lix; cfr. Stoppacci, Silloge, pp. 13-26.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 15

27-Sep-21 8:05:51 PM

16

chapter i

Of particular interest are the four Opuscula, III-VI. Items III and IIII are extracts or sketches which are closely aligned with certain passages of the fourth book of De topicis differentiis, and VI is an overview of themes treated in the same work.21 The corpus of monographs appears first to have been assembled by a certain Martius Novatus Renatus, who commissioned a copy from a student of Priscian in Constantinople named Theodore, who in turn is known to have produced a copy of his master’s Institutiones grammaticae in 526-(?)27.22 Evidence for Renatus’ and Theodore’s engagement with the Boethian corpus takes the form of subscriptions which were later drawn up by an anonymous corrector in the course of collating against the Codex Renati.23 (On the identification of the corrector with a Vivarium copyist named Eusebius, see infra p. 130.) Mistakenly bound in with De divisione in the corrector’s copy was a bifolio drawn from a copy of Boethius’ revised Topics translation, which is otherwise lost. Lorenzo Minio-Paluello’s reconstruction of the foliation of the manuscript for which the index above was originally produced indicates that the Topics fragment formed a part thereof,24 and on the assumptions (a) that the fragment was not interpolated by either Renatus or Theodore and (b) that ancient correctors and editors should not be posited praeter necessitatem, responsibility for the binding error, along with a number of others evidently of the same origin, would appear to lie with the anonymous corrector, working (so I have surmised25) at Vivarium in the second half of the 6th c. — a point which now appears to gain independent confirmation. We will return to this point presently. In an important article Oronzo Pecere has maintained that the four Opuscula (III-VI), like the Liber ante praedicamenta (VII), are not in fact later extracts but authorial drafts of unfinished work that were preserved by Renatus and the corrector, his collaborator, in Ravenna, c. 527-37.26 Now, if the Opuscula are indeed authorial sketches or draft notes rather than extracts later drawn from De topicis differentiis and, in the case of De multifaria praedicatione (V), another source Boethian or otherwise, then it obviously is unnecessary to connect them or the corrector with Vivarium: on this interpretation, the De topicis differentiis interpolations in the ΦΔ-recensions of the Institutiones are simply an unrelated later development. 21 Pecere,

Cassiodoro, pp. 174-176, 203-212; cfr. Stoppacci, Silloge, p. 3, nt. 6. Criticism, pp. 219-222. 23 Magee, Text, pp. 3-4; [ed.] De divisione, p. lviii; Pecere, Cassiodoro, p. 164. 24 Minio-Paluello, Note, p. 104 (364); [ed.] Arist. Lat., V.1-3, p. xxxviii; Magee, Text, pp. 5-6, nt. 11; [ed.] De divisione, pp. lxiii-lxv. 25 Magee, Text, p. 10; [ed.] De divisione, pp. lx-lxi. 26 Pecere, Cassiodoro, pp. 175-176, 183-185. 22 Zetzel,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 16

27-Sep-21 8:05:51 PM

the early textual tradition

17

However, the lack of necessity does not amount eo ipso to an impossibility, and at least two questions appear worthy of further reflection in this connection: Why did Renatus and (or) the (hitherto unidentified) corrector place items III, IIII, and VI after De divisione (II) rather than before or after De topicis differentiis (I), and why did he (they) include item V among them at all? Strictly speaking, the decision between scenarios that presuppose either authorial sketches or later excerpts is not a disjunctive one, for if the Opuscula are remnants of the Boethian Nachlass that were preserved by Renatus and the corrector, as Pecere holds, their precise placement in the index and in the codex for which it was originally devised may nevertheless reflect changes subsequently effected at Vivarium. We will return to this point as well. Priscian and Theodore stand at the centre of two networks of profoundly different cultural resonances. Priscian’s dedication27 of his three minor literary works to Symmachus is reflective of cultural, social, political, and indeed theological bonds that on a broader scale united Roman aristocrats from Ravenna to Constantinople. Renatus, a vir clarissimus et spectabilis, undoubtedly participated in the same network, and his efforts to preserve Boethius’ works posthumously are most readily interpreted as the gesture of pietas performed by a kind of literary executor operating in an environment which was verging on, if not already embroiled in, a violent war that pitted Roman against Goth and Constantinople against Ravenna.28 We may well imagine him organizing the remains of Boethius’ library in Ravenna prior to the advent of war, perhaps even in coordination with Symmachus or other allies of the Anicii, but the only fact of which we can be certain is that he commissioned a copy of the logico-rhetorical monographs from Theodore in Constantinople. Cassiodorus, by contrast, was a latecomer to these developments, who before reaching Constantinople was ignorant even of the fact that Priscian wrote in Latin rather than Greek.29 While in Constantinople, we may further imagine, he was suspiciously viewed by those with past allegiances to the Anicii as a collaborationist who had replaced Boethius as Master of the Offices and then gone on to hold the Prefecture up into the first years of the war. He too strove to preserve Boethius’ works, of course, but his motives cannot have been the same as Renatus’: they may have sprung from much darker feelings and memories. Theodore, a scribe (antiquarius) and palace official (palatinus) of some 27 Gramm.

Lat., III, 405. Troncarelli, Costantinopoli, p. 199; Pecere, Cassiodoro, p. 178. 29 Cassiod., Instit. 2.1.1 (94.1-2 M.): Attico / antico sermone (ΔΦ, drawn to my attention by FT). Cfr. De orthogr. 12 (Gramm. Lat., VII, 207.13-14): Ex Prisciano grammatico, qui nostro tempore Constantinopoli doctor fuit. 28 Cfr.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 17

27-Sep-21 8:05:51 PM

18

chapter i

sort, was obviously acquainted with both Priscian and Renatus, but we cannot assume that he travelled in either of their social circles. There is no evidence to either confirm or refute personal contact with Cassiodorus; the details of his scribal profession are lacking, and assessments vary,30 but it seems reasonable to assume that he produced Renatus’ copy with particular care. With the anonymous corrector (Eusebius) the trail goes cold, although the evidence he provides is indeed the most precious of all. He collated against the Codex Renati at a time when Theodore was still alive and functioning as a civil servant in Constantinople (qui nunc palatinus est), but in the abscence of evidence explicitly connecting him with either Renatus (and Theodore) or Cassiodorus, we are left to weigh probabilities. As indicated above, the interpolated Topics fragment and precise location of the four Opuscula in the corpus of monographs suggest an environment which was lacking in the kind of care that might reasonably have been expected of Renatus and Theodore, which of course in itself does not amount to an argument in support of Cassiodorus and Vivarium. But the traces of De topicis differentiis that survive in both the Opuscula (III-VI) and the ΦΔ-redactions of the Institutiones are unlikely to be coincidental or unrelated phenomena, and since the latter obviously go back to Vivarium, the evidence would appear on balance to indicate that the corrector worked there rather than in Ravenna. The recent emergence of a Carolingian fragment of De divisione preserved in Vercelli has shed further light on these considerations.31 It consists of a folio containing the last section of Boethius’ monograph, including the second half of the interpolated Topics fragment (122a36-b24), and the beginning of the first Opusculum listed in the ancient index (III).32 A collation of its contents reveals the following: shared descent from the common ancestor or archetype of the earliest complete manuscripts of De divisione33; close kinship with the Fleury codex produced under the supervision of Abbo (c. 1000)34; early indications of the contamination fully

30 Cameron,

Pagans, pp. 433-434; Pecere, Cassiodoro, pp. 178-179; Zetzel, Critics, pp. 200, 212. 31 Magrini, Frammento; Codd. Boeth., III.391. The fragment probably dates to the late8th-9th c. (vs. late-9th-10th c.) and bears unmistakable signs of Insular influence. 32 De divis. 889b-892a (vel quando nisi … diligenter expressimus); Comm. spec., PL, 64, 1217c (Quanta sibimet … de singulis). 33 De divis. 890b (46.14 M., aequivocorum vel). 34 Orléans, Mediath., 267 + Paris, BnF, n.a.l. 1611; cfr. Pecere, Cassiodoro, pp. 152-168; Stoppacci, Silloge, pp. 13-17.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 18

27-Sep-21 8:05:52 PM

the early textual tradition

19

visible in the oldest complete copies35; and otherwise unattested variants and errors which, although relatively trivial in themselves, indicate that the manuscript with which the fragment originated was not the direct or indirect ancestor of any of the oldest extant complete witnesses.36 The fragment may be situated as follows in a revised version of the stemma originally designed to illustrate a provisional account of the earliest phases in the history of the transmission of De divisione (and other monographs)37:

c. 505-23, Monographs

Ω (?) Opuscula III-VI

c. 530-35, Codex Renati

c. 560-80, Vivarium

ψ

Topica 122A10-B24, (?) Opuscula III-VI

φ ω

c. 800

(Vercelli fr.)

[?] [?]

c. 1000

De divisione (interpolated)

De divisione (expurgated)

The Vercelli fragment furnishes unique evidence with which to bridge 35 Shared

readings: 889b (44.2 M.) noscitur Vercelli = CGK; c (10) auditor om, cfr. AQ; (21) artem] aut- = AQ; 890a (46.2) cum om = AK; (9) multiplicatis = E; significationem = ACEFGQ; b (15) unius particulae om = AFGKQ; orationum + aut = AFGKQ; (18) similatio = AG; (22) sa. Ro. tr = AE; ut] vel comp = Qc; c (24) singulis ac = G; (26) et2 om = C; (27) ut om = ACEFGQ; d (48.12) est1 di. tr = GK; (14) se partitionibus] separationem = AEKQ; (15) dicimus = AEFQ. 36 Unique readings: 889c (44.7 M.) Troianos1,2] rohn- … -acos ac; d (16) res + res; (17) demonstrandum ac; (21) ait] ut ac; 890a (26) ut] aut pc vid; (46.1) Graecus2 vid; Troiani + troiani; (4) ergti; (6) dicenda + et; (9) dicendum] est praemittit ac; b (18) factum ac; (21) divimus; c (24) aliis] illo ac; (48.3) est1 om ac; d (6) Graecos] romanos ac; (10) ambiguitatem ac; (11) possunt ?ac; (13) geris; (14) tractatum ac; (15) fit ac; 891a (17) dividim; (22) oportet] relinqui praem; (24) addi] ad ac; 892a (28) disiuxit ac; distribuit] destruit ac; (50.2) different ac. 37 Magee, [ed.] De divisione, p. lxv. d

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 19

27-Sep-21 8:05:52 PM

20

chapter i

the gap between Vivarium, c. 560-80, and Fleury and the other monastic centres that gave rise to the oldest surviving copies of De divisione, c. 100050, when the work was already divided between carriers that retained the Topics fragment and others that had been purged of it.38 The fragment cannot, of course, answer a question which has been hinted at earlier, i.e., whether for De divisione and the corpus as a whole there were any other lines of descent reaching back to the later 6th c. Two nodal points stand out in this matrix of interconnections: (a) De topicis differentiis, portions of which made their way into both the Opus­ cula included in the ancient corpus of Boethian monographs and the ΦΔ-redactions of Cassiodorus’ Institutiones; (b) the biographical notices for Symmachus and Boethius, which by way of the Ordo generis in turn link the Δ-redaction of the Institutiones with a Cassiodorian edition of the Consolatio.39 Together they form a basis for speculation about the possibility of a Vivarian Boethii opera omnia of some sort, although (it must be acknowledged) evidence of the sort under consideration here is lacking for the majority of Boethius’ works. His shorter, introductory commentary on De interpretatione, the focus of our concern, is a case in point, in that the oldest manuscripts preserve no subscriptions or collation notes to link the early phases of its transmission with the logico-rhetorical monographs and no symptoms of later excerpting to link it with any redaction(s) of Cassiodorus’ Institutiones. As to the latter, we of course have Cassiodorus’ observation, recorded in the Ω-redaction, that there was a copy of the advanced commentary in six books (sex libris) in the Vivarium library; a reference in the ΦΔ-recensions to a “double commentary” or possibly “two-book disputation” (commenta duplicia, duplici disputatione) indicates however that there was also a copy of the elementary one.40 Otherwise, the oldest surviving manuscripts of the latter contain, as we will see, scattered but clear traces of an ancient edition whose layout recalls that which has been associated with Cassiodorus’ edition of the Consolatio.

38 The

bifurcation at the bottom of the stemma has been modified in order to remove an unintended ambiguity detected by Pecere (Cassiodoro, p. 163). When the Topics fragment was first removed from the text of De divisione is uncertain, so that it is impossible to posit a precise or single point of bifurcation (cfr. Magee, Text, p. 36; [ed.] De divisione, p. lxxiii). 39 Cfr. Troncarelli, Costantinopoli, pp. 191-192; and above, nt. 19. 40 Cassiod., Instit. 2.3.11, 18.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 20

27-Sep-21 8:05:52 PM

CHAPTER II

THE EXTANT TRADITION After Vivarium a mist of obscurity descends upon Boethius’ works, which begins to lift only in the later 8th c. For the 7th and 8th cc., indirect evidence in the form of echoes and citations assists modern efforts to retrace the trajectories cut by individual works, with De arithmetica pointing generally in the direction of Irish computistical and astronomical texts and the Consolatio tending to draw figures such as Aldhelm, Tatwin, and Bede into consideration. Of course, much disappeared along the way,1 and one significant consequence of the losses for us today is the process of triangulation from highly fragmented evidence that is required in order to trace the passage of Boethius’ works through place and time. Manuscripts of the Consolatio illustrate the point that, even where evidence of the sort furnished by manuscripts of the logico-rhetorical monographs is lacking, other paratextual evidence such as the mise-en-page, shifts between script types, scholia, and miniatures can nevertheless shed important light on the lost stages of transmission. The elementary De interpretatione commentary (hereafter c) is a case in point. For that its manuscripts descend from a common ancestor or archetype (hereafter ω) is evident from the numerous points at which they uniformly carry impossible readings2; and that the archetype displayed distinctive paratextual characteristics is made clear by the evidence to which we will turn in sections II.A-B. Manuscript V, the Codex Pagesianus, has rightly been described as the oldest surviving collection of logical texts of the Middle Ages,3 and it may indeed represent our first direct sighting of Aristotle in the medieval West. As viewed in hindsight, however, it cannot rightfully be said to portend the long shadow that Aristotle was destined to cast over medieval philosophy, for the only Aristotelian work included in it is De interpretatione, and that only in the form of the lemmata and citations in c, which do not add up to a complete text of the treatise. As will be pointed out later in the present volume, the manuscript is in two parts. The contents of the first part form a 1 Magee-Marenbon,

Appendix, pp. 308-310; Gruber, Boethius, p. 24. at In Perih. I 109.4 (below, III.B) and 117.2-3 (Magee, Observations, p. 22; utrasque et simul id est [*] adfirmationem in K). 3 Cfr. below, p. 54. 2 Examples

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 21

27-Sep-21 8:05:52 PM

22

chapter ii

coherently arranged logical syllogê which both echoes the systematic order of study laid out by Boethius and other late-ancient commentators (Isagoge – Categories – De interpretatione)4 and foreshadows the medieval Old Logic, with however the Ps.-Augustinian Categoriae decem (Paraphrasis themisti­ ana) and (?) Apuleian Peri hermeneias replacing their Aristotelian counterparts. The Kraków manuscript (K) arranges the same three works identically and, also like V,5 includes excerpts of Alcuin’s De dialectica, whereas C, one of our two Corbie manuscripts (alongside P),6 omits the Porphyry and Ps.-Augustine but includes Aristotle’s De interpretatione, which it situates between the (?) Apuleius and Boethius’ commentary. These three early manuscripts, CKV, are closely related to one another and share textual disturbances that reflect remote contact with some precursor of an interpolated branch of the tradition (GHMO, hereafter φ) which, as we will see (II.C), fills gaps between lemmata with passages excerpted from the continuous translation (hereafter t) or advanced commentary (hereafter C), if not both. Certain individual readings, moreover, point to collation and borrowing that must have occurred before the three manuscripts themselves were produced. At 16b18 (In Perih. I 61.2), for example, Boethius rendered τὰ δὲ τὸν πέριξ with the Latin, illa vero quae circa sunt, which for Ct he revised to (illa vero) quod complectitur,7 a reading incorporated already by early correctors of CK (note especially id est and vel): quae circa sunt] CK : vel praem. Msl : (id est) quod complectitur (Csl)HM : + vel quod complectitur Kmg.

Similarly, at 17b32-33 (In Perih. I 95.9-10) CKV have pulcher and foedus, Boethius’ initial rendering of καλός and αἰσχρός; for C8 and t, however, Boethius revised to probus and turpis, which an early corrector of C obtained through collation and inserted above the line, along with the phrase in which they occur (et est homo … non probus). Conversely, the interpolated witnesses, GHM, intercalate, in the text proper, Boethius’ initial rendering (full phrase) immediately after the revised one, thereby generating a kind of dittography. And, finally, the initial rendering similarly migrated from 4 Boeth.,

In Isag. I 12.17-15.4 B.; In Cat. 161c; Ammon., In Isag. 22.23-23.1 B.; In Cat. 14.24-15.2 B.; In De Int. 4.17-24 B.; Olympiod., Proleg. in Cat. 24.21-25.4 B.; cfr. Cassiod., Instit. 2.3.8-11. 5 Minio-Paluello, [ed.] Arist. Lat., II, p. xlii; below, p. 51. 6 Ganz, Corbie, pp. 64, 150; a third possibly in the fragment, Paris, BnF, lat. 12949 (MinioPaluello, [ed.] Arist. Lat., II, pp. xxx, xlviii). 7 Experimentation with quod complectitur is in evidence at In Perih. I 61.19-25 (cfr. Magee, Composition, p. 16). 8 In Perih. II 169.26-170.5.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 22

27-Sep-21 8:05:52 PM

The Extant Tradition



23

the elementary commentary (c) to the continuous translation (t), as now in Paris, BnF, lat. 12949 (9th c.) and others thereafter.9 The corrections and glossing in CKV reflect dynamic processes of change datable to the early 9th c.; similar disturbances embedded in their uncorrected texts additionally intimate, as we will see (III.C), even earlier strata of change. From these observations we may infer an early phase of editorial activity in the course of which copies of t or (and) C were consulted for purposes of establishing a text of De interpretatione for c, and vice versa. The surmise comports with known evidence for the formation of the composite edition of Boethius’ Latin Categories, three manuscripts of which date to the 9th c., well before the oldest extant witnesses of the continuous translation.10 We may further surmise that the Ps.-Augustinian and (?) Apuleian works were included in KV for reasons other than their mere availability. For the early correctors of CK, as we have seen, certainly had access to Boethius’ continuous translation (included in C) or advanced commentary, if not both, and V commits certain errors which intimate related forms of contact as well; hence the syllogê represented by KV probably reflects a deliberate effort to introduce readers to Aristotle’s notoriously difficult thought through direct exposure only to the lemmata of Boethius’ elementary commentary, which after all is in large part an extended paraphrase intended for beginners. The combined evidence of the composite edition of the Categories and the earliest extant manuscripts of c, in other words, intimates a culture of scholarly engagement with the Aristotelian texts before 800, and indeed, probably before c. 785 (V). The first signals of Aristotle’s rebirth in the Latin West are nearly as fragmentary and indirect as are those associated with his death throes in the later 6th c. In an effort to connect the phases of death and rebirth we turn next to certain distinctive (para)textual features of the manuscripts. The evidence brought under consideration derives from the first book of Boethius’ commentary, comprising De interpretatione 1-9 (16a1-19b4), and is of three types. II.A. Subject Headings The manuscripts assign titles or subject headings to what in modern editions figure as cc. 2-4 of De interpretatione. That is consistent with Ct, Ammonius’ commentary,11 and, more generally, with subject headings and 9 Arist.

Lat., II, app. crit. at 11.16 (note also at 13.17, 16.11). Lat., I.1-5, p. 44 (Lg, Re, Vc); cfr. Asztalos, Transmitter, pp. 371-372. 11 In Perih. II app. crit. at 52.28, 65.29; Arist. Lat., II, 6.4, 7.1, 20; Ammon., In De Int., app. crit. at 29.29, 47.6, 58.4, 81.3 B. 10 Arist.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 23

27-Sep-21 8:05:52 PM

24

chapter ii

titles preserved in certain manuscripts of De divisione and the Consolatio12; moreover, before cc. 10 and 14 (19b5, 23a27) some manuscripts of t echo the section divisions (κεφάλαια, τμήματα) adopted or devised by Ammonius for his commentary.13 Titles associated with the lemmata introducing cc. 2-4 appear as follows in our manuscripts of c (cfr. III for editorial conventions): 2. 16a19-20 (In Perih. I 45.30-46.2) DE NOMINE nomen … separata BOETIUS CDKLPV | DE NOMINE (ARISTOTELES) nomen (Cc)FGM : INCIPIT DE NOMINE nomen H : DE DEFINITIONE nomen O | separata] F : + BOETIUS Fc 3. 16b6-7 (In Perih. I 55.21-22) verbum … significat FVmg : EXPLICIT DE NOMINE INCIPIT DE VERBO CKLPV : EXPLICIT DE NOMINE INCIPIT DE VERBO verbum est … significationem tenens (61.26-62.1) Lc | verbum M : EXPLICIT DE NOMINE INCIPIT DE VERBO verbum CcDH : HISTORIA INCIPIT DE VERBO verbum G : DE VERBO verbum Omg | extra + DE VERBO M | significat] significativa est … significationem tenens (61.28-62.1) H : + et est … nota (16b7-8) M 4. 16b26-28 (In Perih. I 66.26-28) oratio … adfirmatio CMV | DE ORATIONE (HISTORIA) oratio Cc(mg)DF(G)LmgOP : [**********] oratio K : EXPLICIT DE VERBO INCIPIT DE ORATIONE oratio H | adfirmatio + BOETIUS Cc

Immediately before the first lemma of c. 2 the archetype (ω) had the subject heading, De nomine, as attested by all but the patently confused O. To judge from CDFcKLPV, it also supplied Boetius after the lemma to mark the transition from Aristotelian lemma to Boethian commentary; the four exceptions, GHMO, are interpolated witnesses (see further under II.C, below). Hence the archetype probably read as follows here: (…) cursum intendit. De nomine. Nomen ergo est vox significativa secundum placitum sine tempore, cuius nulla pars est significativa separata. Boetius. Omnis definitio (etc.). We will turn to the question of majuscule and minuscule scripts presently. Of primary interest at c. 3 is the competition between subject heading and lemma. For prior to correction, CKLPV have a subject heading but no lemma, whereas F has a lemma but no subject heading, and DGHMO variously have both. Confusion abounds. The corrector of L reaches for a passage of commentary in a desperate effort to supply the missing lemma, and nearly the same commentary passage resurfaces as part of the lemma in H. M’s lemma is infiltrated by De verbo and supplemented by the directly 12 Magee,

Text, p. 11; Troncarelli, Tradizioni, pp. 1, 57-59. Lat., II, p. xliii (cfr. Magee, Composition, p. 24); Ammon, In De Int. 7.15-18, 86.26, 159.24, 214.8-9 (with app. crit. at 1), 251.9 (app. crit.) B. 13 Arist.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 24

27-Sep-21 8:05:52 PM



The Extant Tradition

25

ensuing phrase of De interpretatione, anticipating a phenomenon to which we will return presently (II.C). CV begin without the lemma then move independently in the course of correction, for although both acquire the lemma, V apparently abandons the subject heading in the process. There is, as we will see, evidence to indicate CV’s shared descent from an exemplar (hereafter δ) which contained a number of double readings; here, as in many other passages, C eventually hits the mark in assessing the options offered by δ+δc, whereas V apparently reads them disjunctively, substituting one omission for another. Like M, O reduces the title to De verbo, while G’s distillation preserves the incipit also found in CDHKLPV. In all probability the subject heading in the archetype here took the form, Explicit de nomine, incipit de verbo, and was not supplemented by Boetius to indicate the transition from lemma to commentary. The archetype may well have omitted the lemma and then acquired it through a correction whose precise import was subject to misinterpretation. This possibility receives support from a number of passages in which confusion associated with the copying of lemmata had a disruptive effect on the directly related passages of commentary. Three cases will serve to illustrate the point. First, at In Perih. I 48.6-7 Boethius wrote haec (hoc in LO) autem propositis docet exemplis to introduce the next lemma (16a21-22). The transitional phrase was evidently omitted then restored after the lemma owing to a correction that antedated (Cc)DL: haec … exemplis FGOP : om. CHKLcMV : post ferus (16a22) tr. CcDL

Words cited from the previous lemma at In Perih. I 48.4-6 (nomen … secundum placitum … sine tempore, cuius, 16a19-20) may have been a contributing factor to the omission. For the manuscripts contain evidence to suggest that the archetype had citations, subject headings, and lemmata which, at least for Boethius’ commentary on the first chapters of De in­ terpretatione, were written in a majuscule script; and if the citation at In Perih. I 48.4-6 was a case in point, then it would have been natural for scribes to leap inadvertently to the next lemma and so omit the transitional phrase. In P, we may note, both the transitional phrase and lemma are in Rustic Capitals, while D employs Capitals for the first half of the lemma and minuscules for the second; F, by contrast, has the whole of the lemma in Uncials (slightly hybridized), while MV have only the first two words (in nomine) in either Capitals (M) or Uncials (V). Again, at In Perih. I 49.3 Boethius shifted from commentary to lemma (16a22-26) without providing a transitional phrase to alert readers to the change. FGO, however, import the formula, dicit enim, employed elsewhere in the commentary to mark

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 25

27-Sep-21 8:05:52 PM

26

chapter ii

approaching lemmata, and here too manuscripts waver between scripts, although without thereby disturbing the immediately surrounding portions of commentary: FK have the whole lemma in Uncials, while P has it in Capitals; L has the first third in Capitals, while V wavers between minuscules and a degraded form of Uncials. Finally, the lemma, 16b21-23 (In Perih. I 64.13-14), consists of two phrases, (a) sed si est vel non est nondum signifi­ cat and (b) neque enim ‘esse’ signum est rei vel ‘non esse’, which Boethius cites in full after ten lines of exegesis (In Perih. I 64.23-25). Patterns of oscillation between majuscules and minuscules are most pronounced in DLP: Lemma Citation

(a) sed si … nondum significat DL minuscules; P Capitals DLP minuscules

(b) neque enim …‘non esse’ DL minuscules; P Capitals DLP Capitals.

Related patterns are discernible also in certain other symptoms. Thus K mirrors P in rubricating the whole lemma but only citation-(b), H rubricates only lemma-(b), which M omits, and so on. These data suggest that in an early ancestor both the lemma and citation were written in a majuscule script which became variously altered in the course of repeated copying, and M illustrates an eventual loss of text stemming probably from growing confusion over what was finally to be counted as lemma or commentary (citation). Some such confusion would appear to underlie the omissions and dislocations in evidence at the beginning of c. 3. As to c. 4, the phrase, Explicit de verbo, incipit, in H would appear, like the Incipit at c. 2, to be a case of hypercorrection, bringing the subject heading into formal alignment with the one at the head of c. 3. The corrections to C intimate another duplex lectio in δ, the parent of CV; and although K, their next of kin, sides with V in omitting the subject heading, the lacuna before oratio breeds suspicions about the exemplar (hereafter μ) from which both it and δ appear to descend. The subject heading here evidently took the form, De oratione, in the archetype. It may initially have been omitted from μ then reinstated through correction; evidence for an indicator (Boetius) of the transition from lemma to commentary is negligible. See further under III.A, below. These passages are highly revealing. The three subject headings De no­ mine, Explicit de nomine, incipit de verbo, and De oratione would appear to date back to the archetype, which in this respect conformed with Ct and the Greek tradition as represented by certain manuscripts of Ammonius’ commentary. Cassiodorus’ reorganization (Instit. 2.3.11) of the six topics listed at De interpretatione 16a1-2 is almost certainly of interest here as well: (1) de nomine, (2) de verbo, (6) de oratione, (5) de enuntiatione, (4) de

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 26

27-Sep-21 8:05:52 PM

The Extant Tradition



27

adfirmatione, (3) de negatione, de contradictione. Whether or not the subject headings reflect Boethius’ intention remains an open question. The interplay of scripts suggests a hierarchy thereof in the archetype, perhaps in the manner of Vat. lat. 3363, an early manuscript of the Consolatio in which the titles are written in Rustic Capitals, the poetry in Uncials, and the prose sections in minuscules.14 By the last quarter of the 8th c. or earlier the three subject headings had become unstable due to collation and correction, leading to the fluidity of our oldest extant witnesses. II.B. The Framing of Lemmata and Citations As we have seen, the manuscripts preserve traces of signs or editorial aids for distinguishing between Aristotelian text, in the form of lemmata and citations, and Boethian commentary. In fact, different systems of demarcation are in play, most of which fade out in the third chapter of De interpretatione. For cc. 1-3 Boethius drew up twenty-three lemmata, which in the manuscripts are framed as follows: 16a1-2 (In Perih. I 35.4-6) primum … oratio HLMO | primum F : INCIPIT HISTORIA primum CP : FINIT PRAEFATIO INCIPIT TEXTUS primum D : [***40***] primum KV : //// primum G | oratio] C : + EXPOSITIO CcDFP 16a3-8 (In Perih. I 36.22-28) sunt … eaedem DFGHMO | HISTORIA sunt CmgP | [*******] quorum … eaedem litteris minutis, post capitaneas, exarata L | eaedem] hae (cfr. 40.28, II 39.33) P : + EXPOSITIO CKV : + EXPLICIUNT MODI LOCUTIONUM L 16a8-9 (In Perih. I 41.9-11) de his … negotii] + BOETIUS Gmg 16a9-13 (In Perih. I 41.16-21) est autem … veritasque HMO | est autem] C : ARISTOTELES praem. Cc | veritasque + BOETIUS (PRINCIPIUM) C(D) FGK(LP(mg))V 16a13-16 (In Perih. I 43.15-18) nomina … falsum est HKMOV | ARISTOTELES nomina C : PRINCIPIUM nomina DL : SEQUITUR nomina G | falsum est] C : + BOETIUS CcDFGLmgP 16a16-18 (In Perih. I 44.19-22) huius … tempus] C : ARISTOTELES huius … tempus BOETIUS Cc | PRINCIPIUM praem. DLmg | + BOETII HISTORIA Gmg 16a19-20 (In Perih. I 45.30-46.2) (see under II.A, above) 16a21-22 (In Perih. I 48.8-10) in nomine … equus ferus] + BOETIUS CFPmg : + HISTORIA Gmg 16a22-26 (In Perih. I 49.4-8) at vero … ferus] C : ARISTOTELES at vero … ferus BOETIUS Cc | nominibus] nomini BOETIUS Kcomp | ferus + BOETIUS DmgFPmg 16a27-29 (In Perih. I 50.9-12) secundum … nomen] ARISTOTELES secundum … 14 Codd.

Boeth., III.547; Troncarelli, Cogitatio, p. 244, with pp. 61-62.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 27

27-Sep-21 8:05:52 PM

28

chapter ii

nomen BOETIUS C | + BOETIUS DmgFPmg 16a30-32 (In Perih. I 51.20-23) non homo … infinitum HKMOV | non homo] C : ARISTOTELES praem. Cc : HISTORIA praem. Gmg | infinitum] Lc : + BOETIUS CDFLP 16a33-b1 (In Perih. I 53.5-6) Catonis … nominis] C : ARISTOTELES Catonis … nominis BOETIUS Cc | SEQUITUR (HISTORIA) praem. (Gmg)P | + PRINCIPIUM Dmg 16b1-5 (In Perih. I 54.11-16) ratio … mentitur CHKMOV : ARISTOTELES ratio … mentitur BOETIUS Cc | HISTORIA ratio ?Gmg | mentitur + BOETIUS F : + PRINCIPIUM L : + PRINCIPIUM BOETIUS D(mg)P 16b6-7 (In Perih. I 55.21-22) (see under II.A, above) 16b7-8 (In Perih. I 56.14-15) et est … nota] + PRINCIPIUM DP 16b8-10 (In Perih. I 57.3-5) dico … esse] C : ARISTOTELES dico … esse BOETIUS Cc | HISTORIA praem. Gmg | esse + PRINCIPIUM DP : esse […] F 16b10-11 (In Perih. I 57.23-25) et semper … in subiecto CDFKMOV | HISTORIA et semper Gmg | in subiecto + BOETIUS Cc : + LEGE DILIGENTER DE SUBIECTO FmgHLP 16b12-14 (In Perih. I 59.4-7) non currit… verbum] ARISTOTELES non currit … verbum BOETIUS C  | PRINCIPIUM praem. DL : HISTORIA praem. Gmg 16b15 (In Perih. I 60.3-4) quoniam … non est] C : ARISTOTELES quoniam … non est BOETIUS Cc  | HISTORIA praem. Gmg  | non est + PRINCIPIUM DL : non est [***15***] HM 16b16-18 (In Perih. I 60.27-61.2) similiter … circa sunt] C : ARISTOTELES similiter … circa sunt BOETIUS Cc | PRINCIPIUM praem. L | + HISTORIA Gmg 16b19-22 (In Perih. I 62.5-8) ipsa … significat] ARISTOTELES ipsa … significat BOETIUS C | III praem. ?F : HISTORIA praem. Gmg 16b21-23 (In Perih. I 64.13-14) sed si … ‘non esse’] C : ARISTOTELES sed si … non esse BOETIUS Cc   | HISTORIA praem. Gmg 16b23-25 (In Perih. I 65.9-12) nec si … intellegere] C : ARISTOTELES nec si … intellegere BOETIUS Cc | HISTORIA praem. Gmg

Moreover, six citations adduced in support of particular points of exegesis on cc. 1-2 are similarly framed: In Perih. I 39.2-4 (16a5-6) et quemadmodum … eaedem voces FGHKcMO  | HISTORIA et quemadmodum … eaedem voces EXPOSITIO C(mg)LP | TEXTUS et quemadmodum D | eaedem voces + EXPOSITIO KV In Perih. I 39.12-15 (16a6-8) quorum … eaedem DHMO | HISTORIA quorum … eaedem BOETIUS C(mg)LP : HISTORIA ARISTOTELES quorum Cc | eaedem + BOETIUS GKV : eaedem // F In Perih. I 40.7-9 (16a7-8) et quorum … eaedem] HISTORIA praem. CmgLP :

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 28

27-Sep-21 8:05:53 PM



The Extant Tradition

29

HISTORIA ARISTOTELES praem. Cc In Perih. I 40.26-28 (16a6-8) quorum … hae HMO | HISTORIA quorum … hae BOETIUS C(c)LP | hae + BOETIUS DFG mgKV In Perih. I 42.25-26 (16a12-13) circa … falsitasque DFHMO | circa] C : HISTORIA praem. Cc(mg) | falsitasque + BOETIUS CGKLPV In Perih. I 52.31-53.1 (16a30-31) at vero … appellari HKMV | at vero] CFcGLc  : ARISTOTELES praem. Cc : SEQUITUR (VOCABULUM) praem. (D(mg)) FGcLO(P) | appellari + BOETIUS C

The combined evidence of these passages may be grouped under the following five headings for purposes of analysis: 1. (a) lemma, citation + LEGE DILIGENTER DE SUBIECTO; (b) TEXTUS + lemma, citation; (c) lemma, citation + EXPLICIUNT MODI LOCUTIONUM. 2.

SEQUITUR (VOCABULUM) + lemma, citation (+ BOETIUS).

3.

ARISTOTELES + lemma, citation.

4. (a) HISTORIA + lemma, citation (+ EXPOSITIO); (b) lemma, citation + (BOETII) HISTORIA. 5. (a) PRINCIPIUM + lemma, citation; (b) lemma, citation + PRINCIPIUM (BOETIUS PRINCEPS).

(1) is a heterogeneous collection variously attested by DFcHLP. Under (a) appears the interpolated editorial note, lege diligenter de subiecto. LP write the lemma in Rustic Capitals, appending the interpolation, also in Capitals, as though it were a part thereof (16b10-11): ET SEMPER EO­ RUM QUAE DE ALTERO DICUNTUR NOTA EST, UT EORUM QUAE DE SUBIECTO VEL IN SUBIECTO LEGE DILIGENTER DE SUBIECTO. In H, it is rubricated as though forming part of the lemma (which it fragments), whereas in F it is a marginal note written in an Insular minuscule. The phrase, ἢ ἐν ὑποκειμένῳ (vel in subiecto), dates back at least to Porphyry15 but is rightly athetized by Minio-Paluello and Weidemann; the lege diligen­ ter in FcHLP is an editorial note affirming de (subiecto) over in evidently interpreted as a false variant (vel) or scholium. How these manuscripts acquired the note is uncertain, but it may be noted that H exhibits a number of similar points of interest, one of which we will come to presently (III.B). Textus (b) appears only in D, first as part of a longer phrase introducing 16a1-2, and then to announce a citation at In Perih. I 39.2-4 (see further 15 Ammon., In De int. 50.7-14 B. Boethius regards vel in subiecto as authentic (In Perih. II 68.24-69.22).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 29

27-Sep-21 8:05:53 PM

30

chapter ii

[4], below). Finally, expliciunt modi locutionum (c) is restricted to L (16a38), where it brings Aristotle’s remarks on written, spoken, and mental linguistic “modes” to a conclusion; it appears to be an editorial gloss. Neither (b) nor (c) furnishes evidence useful for filiation, and there is no reason to believe that either reaches far back beyond either D or L. The sequitur under (2) poses another question of authenticity. It introduces the lemma, 16b7-8, in a way that led Meiser to misconstrue the latter as a citation (In Perih. I 56.14-15): (…) esse non possit. sequitur : ‘et est semper’ (etc.). Here sequitur is attested by all but H (expunged in LV) and written in majuscules by all but GMO. In fact, KLPV have both sequitur and the lemma in majuscules (F rubricates both), whereas CD have sequi­ tur in majuscules but the lemma in minuscules; DP, moreover, append principium (in majuscules) to the lemma (see [5], below). It is certain that Boethius penned the transitional sequitur, which in the archetype (ω) appears to have been misinterpreted as part of the lemma and so copied in majuscules. Doubts concerning the relationship of sequitur to the lemma gradually set in, with the telling results that in CD the vestigial majuscules attached to sequitur rather than the lemma and that the interpolated witnesses, GHMO, eschewed majuscules and, in the case of H, actually rejected sequitur. At In Perih. I 52.30-31, by contrast, sequitur (vocabulum) introduces a citation (16a30-31) in a manner that patently betrays an editorial interpolation: (…) quod ipse testatur dicens SEQUITUR ‘at vero’ (etc.). Here DLP have the intrusive sequitur in majuscules but the citation itself in minuscules. The mobility of sequitur is perhaps best illustrated by its blunt intrusion, in G, upon the sentence leading to a citation at In Perih. I 42.2526 (16a12-13): (…) retinet significationem SEQUITUR docuit autem per hoc quod ait ‘circa compositionem’ (etc.). This sequitur may have inhabited the margins of an earlier copy before migrating to the text proper in G, where it is abbreviated S (as also before 16a13-16); it evidently reflects a confused inference drawn from one of the traditional marginal indicators for lemmata and citations: s ~ S(equitur). Where sequitur interrupts Boethius’ syntax it is obviously spurious, a Boethian usage misapplied by a later editor; the fluctuations in script suggest that at least some of the editorial confusion associated with this transitional formula originated at a relatively early point in the history of transmission. A number of the lemmata and citations introduced by sequitur or other transitional formulae are supplemented by Boetius (abbreviated B,16 with medial points and titulus) to mark the transition back to commentary.17 To judge from the support 16 Occasionally 17 See

BO, BB. further under II.A, above.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 30

27-Sep-21 8:05:53 PM



The Extant Tradition

31

it receives from CDFGKLPV at 16a9-13 (In Perih. I 41.16-21) in particular, this use of Boetius is ancient and probably related to an interlocutor’s sign (B) preserved in some manuscripts of the Consolatio (e.g., Kraków, Bibl. Jagiell., Berol. Lat. Qu. 939); although surviving in G, it is conspicuously absent from the other interpolated witnesses (HMO). (3) The use of Aristoteles (abbreviated A, ARIS) as a transitional formula is unique to C (Cc), appearing first in connection with citations (In Perih. I 39.12-15, 40.7-9) and then with a lemma (16a9-13). The framing formula, Aristoteles (lemma, citation) Boetius (commentary), is stubbornly pursued by C from 16a9 on, after brief flirtation with historia … expositio (see [4], below). It seems clear that C reflects an editorial attempt to rationalize or update what must have appeared an archaic framing formula, and the results recall certain early Carolingian attempts to replace older, arcane systems of signs for speakers in manuscripts of the comedies of Terence. That C experimented in the course of copying and correcting is evident from the vacillating pattern: historia (16a1-8); historia Aristoteles (In Perih. I 39.1240.9 [16a6-8]); Aristoteles (16a9ff.); historia (In Perih. I 40.26-42.26 [16a613]); Aristoteles (In Perih. I 52.31-53.1 [16a30-31]). Aristoteles (… Boetius) appears to be an innovation devised at Corbie in the first half of the 9th c. (4) Most noteworthy about the transitional formula, historia (abbreviated HIS), is the wedge it drives between the passages in which it introduces (a) lemmata or citations and those in which it introduces (b) commentary — in effect, between philosophical (Aristotle) and exegetical (Boethius) notions of “inquiry.” Although the attestation for these chapters is restricted to CGLP, it should be noted that historia is also found in D, where (e.g.) it intrudes upon the approach to a paraphrase (In Perih. I 132.24 [19b2224]): (…) nominatur HISTORIA ergo nunc hoc dicit (etc., as also in P). Some intriguing patterns of filiation are in play. For in CLP historia ushers in the first two lemmata and a series of citations (16a1-8; In Perih. I 39.242.26), whereas G explicitly disambiguates in favor of commentary (Boetii historia) on first usage (16a16-18, 21-22) but uses it to introduce lemmata from 16a30 on (an exception at 16b16-18). In LP historia intrudes upon the approach to a citation at In Perih. I 38.25-26 (16a4): (…) mentisque conceptio HISTORIA at vero quod addidit his, ‘et ea quae’ (etc.); D inserts textus (see [1], above) at precisely the same point. This would appear to be another case of confused migration from margin to text at an earlier stage in the history of transmission. Insofar as the notion of “inquiry” is in principle equally applicable to Aristotle’s philosophy and Boethius’ explication thereof, the confusion experienced by scribes who encountered historia in the exemplars from which they copied is understandable; it seems reasonable to assume, however, that the originator of this transitional formula

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 31

27-Sep-21 8:05:53 PM

32

chapter ii

intended it for either Aristotelian text or Boethian commentary but not both. Patristic usage, we may recall, includes good evidence for the use of ἱστορία to indicate the literal or historical as opposed to more abstract and theoretical senses of Scripture (Lampe, Patristic Greek Lexicon, q.v., C.3), an observation which may help to elucidate the sense of historia in play here. Among philosophical commentators of the later 5th and 6th cc. lemmatized commentary organized according to general analysis (θεωρία) supported by the explication of text (λέξις) was an accepted part of the landscape. As we will see below (II.C), however, Boethius’ commentary begins with just such a mode of exegesis but gradually develops into a blended genre favoring philosophical paraphrase, which in its privileging of clarity and simplicity over problem-based analysis is particularly well suited to the needs of students who are approaching Aristotle’s notoriously rebarbative and elliptical prose for the first time. It seems natural to surmise that historia was originally applied in consideration of the fact that Boethius’ elementary commentary is essentially a “literal reading” or “narration” of Aristotle’s text, not a higher-order exploration thereof as in the case of the advanced commentary; if correct (the evidence of the advanced commentary has yet to be assessed), this surmise would suggest that historia was meant to signal commentary rather than lemmata or citations. Although the scribal confusion alone is probably an indication of considerable antiquity, the origins of this use of historia remain obscure. Immediately after the first two lemmata (16a1-8) and an early citation (In Perih. I 39.2-4) the word, expositio (abbreviated EXP), is used to mark — for scribes, far more transparently than historia did — the transition from text to commentary. It is variously attested by CDFKLPV and is almost certainly of ancient origin. The abbreviation, EXP, we may note, appears in a 6th-c. manuscript of Vivarian origin (St. Petersburg, Nat. Bibl., Q.v.I., 6-10, f. 61r [CLA, XI.1614]), where it may stand for expositio misinterpreted as explicit; and an abbreviation of similar import, INTERP(retatio), serves to mark legal glosses in a 6th-c. copy of the Breviarium Alarici (Munich, Bayer. Staatsbibl., clm 22501, f. 44r [CLA, IX.1324]). (5) poses a similar interpretive challenge. To judge from two appositional formulations discussed immediately below, princeps18 would appear to be the resolution of the suspended form in which it most often occurs (PRIN19); and if so, then the transitional formula may originally have been intended to signal the notion of “author” or “master,” i.e., the philosopher 18 Possibly principalis (substantive)? Cfr. In Perih. II 193.26-27 (Peripatetici enim, quorum Aristoteles princeps est). 19 Occasionally PR, PRN, also PIRN and RIN (sic).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 32

27-Sep-21 8:05:53 PM



The Extant Tradition

33

Aristotle rather than his commentator Boethius. Among the passages under consideration here the formula is confined to DLP, whose scribes evince pointed confusion in signaling both (a) Aristotelian text, when placing it before lemmata (16a13-18, b12-14, 16-18), and (b) Boethian commentary, when placing it after them (16a9-13, 33-b5, 7-10, 15), with two appositional formulations included among the latter (Boetius princeps, princeps Boe­ tius, 16a9-13, b1-5). Independent evidence may however point to princip­ ium (“primary text, what gives rise to commentary”) as the true resolution of PRIN. For in the margins of BAV, Pal. lat. 271 (fols. 116v, 173v, 177v), a 9th-c. copy of Cassiodorus’ Expositio in Psalmos, there appear alongside certain editorial aids that are known to have originated with Cassiodorus the abbreviations, PR and PN. They are written in a script that evokes a late-antique Uncial, and mark Biblical citations for which no author as such can have been intended; they additionally recall PP, a formula devised by Cassiodorus to insulate “protected” sacred text (lex divina) from contamination by surrounding commentary (Instit. 1.26.2). Hence if PRIN too is of Vivarian origin, then it almost certainly stands for principium, signaling (Aristotelian) text. The framing formulae (textus, sequitur) and editorial insertions grouped under (1) and (2) appear to be of early medieval rather than late-antique origin. C evidently strikes out on its own in settling on Aristoteles … Boetius (3) as the most readily intelligible framing formula. Historia and princip­ ium (4, 5) proved sources of confusion for scribes seeking to distinguish between lemmata and commentary (or citations); like expositio and Boeti­ us, both would appear to be of ancient origin. Finally, we may recall legislation of 553 restricting the propagation of unauthorized interpretations of Scripture (Nov. 146 praef.), and the pronounced tendency of our three latest — and interpolated — witnesses (HMO) to eschew framing formulae: a conservative impulse of the age of Justinian that undoubtedly favored the development of editorial aids such as principium and historia, on the one hand, and scholastic requirements of the medieval schools, on the other. II.C. The Interpolated Tradition We turn finally to the most distinctive trait of the textual tradition, the interpolated branch (φ), which was first detected by Meiser20 and is represented by GHMO, as noted above. Roughly midway through De interpre­ tatione, c. 7 Boethius abruptly shifts to a blended mode of commentary in 20 Meiser, [ed.] In Perih. I, pp. viii-ix; cfr. Minio-Paluello, [ed.] Arist. Lat., II, pp. xlviii-l, lii-liii; Magee, Observations, pp. 18-25.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 33

27-Sep-21 8:05:53 PM

34

chapter ii

which reduced lemmata21 are supplemented by embedded citations and paraphrases. The change takes place at In Perih. I 93.3-25: 7. 17b20 contrarie vero universalem adfirmationem et 21 universalem negationem

93.5 Planissime omnes exsequitur dicitque contrarias 6 universalem adfirmationem et universalem negationem. 7 has enim nos quoque supra descripsimus, hic vero 8 nunc easdem contrarie demonstrat opponi. illud quoque 9 addidit, quod eas impossibile sit in eodem veras 10 aliquando cognosci. nam sicut contrariorum natura 11 in eodem non potest inveniri, neque enim uno eodemque 12 tempore aliquid nigrum est atque album, sic 13 quoque nec contrariae ut utraeque simul sint verae 14 fieri potest. quod autem adiecit, his vero oppositas 15 contingit in eodem, particularem adfirmationem 16 et particularem negationem designat, particularis namque 17 adfirmatio universali negationi opposita est contradictorie, 18 particularis vero negatio universali adfirmationi. 19 contrariarum igitur oppositae possunt in 20 eodem verae aliquotiens inveniri, id est particularis 21 adfirmatio et particularis negatio verae ut sint in 22 aliquibus fieri potest, ut est ‘quidam homo albus 23 est’, ‘quidam homo albus non est’: utraeque sunt 24 verae. 7. 17b26 Quaecumque igitur contradictiones …22

Words and phrases (here capitalized) drawn from the non-lemmatized portion of Aristotle’s text (ut … albus, 17b21-26) have been here woven into the syntax of the commentary in a manner reminiscent of Themistius’ philosophical paraphrases, some of which Boethius may well have known23; Aristotle’s ipsissima verba are flagged at two points (8-9, 14 = 17b23-24) but otherwise silently absorbed into the exegesis. Although the new mode of exposition undoubtedly reflects an effort to clarify Aristotle’s notoriously obscure style, there is no obvious reason for its appearance at this point in the commentary; once underway, however, it moves more smoothly and rapidly than the θεωρία καὶ λέξις mode with which Boethius began the commentary (tota autem ratio sensus … atque hoc est quod ait, In Perih. I 37.438.15). It is as though we are here granted privileged access to Boethius’ methodological reflection at the first moment of its occurrence to him, a 21 Scribes

occasionally mistook citations for lemmata and interpolated or otherwise edited accordingly. The difference between the two is not always obvious. 18a23-25 (In Perih. I 102.17-20), for example, is introduced as a lemma (hoc est enim quod ait) but is in fact an embedded citation; hence, id est ea quae dicit ‘est tunica alba’ (18a20), an explanatory aside omitted by GacHM. 22 Critical apparatus (minus the readings of K) and comments in Magee, Observations, pp. 18-20. 23 References to Themistius in Top. diff. (pass.) and at In Cat. 162a and In Perih. II 4.2-3.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 34

27-Sep-21 8:05:53 PM



The Extant Tradition

35

possibility which may receive support from the fact that he immediately reverts to fully lemmatized commentary (In Perih. I 93.25-97.19, 17b26-37) before effectively24 settling on the blended mode of exposition (In Perih. I 97.20, 17b38). The most significant consequence of the change in exegetical method is that, on finding themselves confronted with a commentary whose lemmata accounted for only two thirds of Aristotle’s treatise, some medieval editors or scribes restored the “lost” passages by excerpting from t or (and) C — a process reminiscent of, if not actually related to, that which gave rise to the composite edition of Boethius’ Categories translation. Traces of their editorial handiwork are evident in details such as the use of vel to mark variant readings and of usque (sc., ad) to indicate the extent of contracted interpolations.25 At one point the editor of (or behind) H explicitly remarks upon the extent of a particular interpolation.26 And a glimpse into what appears to have been a related process is offered by Paris, BnF, lat. 12949, a 9th-c. fragment, possibly from Corbie,27 which aggressively excerpts from the final pages of the commentary (sic opinio … contraria suspicari, In Perih. I 222.6-225.11) in an effort to fit the selected passages into a single side (f. 81r). Although it is both convenient and meaningful to speak in terms of interpolated (φ) and non-interpolated (hereafter ς) branches of the tradition, two cautionary notes must be sounded. First, the ς-witnesses (CDFKLPV) evince signs of collation and editorial intervention similar to those seen in the φ-witnesses (GHMO), at least five examples of which have been encountered above: nomini Boetius (abbrev., medial points, titulus) for nominibus in K at 16a22-26 (In Perih. I 49.4-8; II.B); subject heading displaces lemma in CKLPV, commentary infiltrates lemma in Lc at 16b6-7 (In Perih. I 55.2122; II.A); importation of quod complectitur in CslKmg at 16b18 (In Perih. I 61.2), and of probus … turpis in Cc at 17b32-33 (In Perih. I 95.9-10; II). Evidence for the type of scholarly activity that led to the φ-interpolations, in other words, is already present in the ς-branch, and the contamination resulting from that activity has a disruptive effect on the line separating the two families. Second, the interpolation of passages from t (C) is not in fact restricted to commentary after In Perih. I 93.3 (17b20), for at least four such interpolations appear earlier: 24 There are occasional reversions to continuous lemmata and (or) citations after In Perih. I 97.20. 25 The latter in G at In Perih. I 93.4 (17b21-26, see Magee, Observations, p. 19) and 98.24 (18a3-12, si autem aliud aliquid vel de alio idem, non opposita, sed erit ab ea diversa usque et quando vera vel falsa).  Cfr. Minio-Paluello, [ed.] Arist. Lat., II, p. lii. 26 See on III.B, below. 27 See above, ntt. 6, 9.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 35

27-Sep-21 8:05:53 PM

36

chapter ii

16b7 (In Perih. I 55.22) significat + et est … nota (16b7-8) M 16b30 (In Perih. I 68.2) addatur + sed non … syllaba (16b30-31) M 17a5 (In Perih. I 70.27) falsa] L : + enuntiativa est ubi veritas et falsitas adest HLc(bis) 17b15 (In Perih. I 91.9) praedicetur + ut omnis homo omne animal (est H) (17b15-16) GHM

16b7 has been discussed above (II.A). After 16b30, M adds a phrase from 16b30-31 despite the fact that the latter is cited at In Perih. I 68.29 and provides the bridge to the next lemma (16b31-33, In Perih. I 69.8-11). Of the ς-witnesses, C eventually identifies the citation and inserts its usual markers for Aristotle and Boethius (A, B), while F through rubrication extends it to non est (In Perih. I 69.1). 17a5 attracts a gloss. In H, the interpolation follows as though it were commentary, while L reflects pronounced confusion over both lemma and commentary (In Perih. I 70.24-71.1): [****] vera neque falsa, Enuntiativa est ubi veritas et falsitas adest, ENUN­ TIATIVA VERO NON OMNIS SED IN QUA VERUM VEL FALSUM INEST, Non autem in omnibus, ut deprecatio oratio quidem est sed neque ORationis ut supra iam diximus, multae sunt species enuntiativa est ubi veritas et falsitas adest Est enim …

The underlined phrases represent the full lemma (17a2-5), which is written in a bizarre combination of majuscules and minuscules and dislocates vera neque falsa prior to correction. The interpolation (italicized) appears twice, originally in the commentary (sunt species … Est enim) and then (later hand) in the lemma (neque falsa … ENUNTIATIVA VERO). From L we may infer several antecedent levels of activity, including collation, correction, and glossing; and to connect HL at least two editorial steps must be postulated, restoration of the disintegrated lemma and a repositioning of the interpolation. Finally, after 17b15 GHM add a phrase from 17b15-16, which however is otiose given the citation at In Perih. I 91.22, which furnishes the bridge to the next lemma (17b16-20, In Perih. I 91.26-92.4). This passage marks the first unmistakable appearance of the φ-interpolator.28 Listed below are the interpolated lemmata (citations) in the first book of c. The second column indicates their fullest extension, and the third their carriers arranged according to patterns of agreement; the broken line represents the point after which G is unavailable for consultation: 28 Cfr.

Magee, Observations, p. 20.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 36

27-Sep-21 8:05:53 PM



37

The Extant Tradition

17b20-21 (In Perih. I 93.3-4)

17b26

(G)M – HO29

17b24 (In Perih. I 93.14-15)

17b26

H

17b38-40 (In Perih. I 97.20-23)

18a1

(G)HMO

18a3-4 (In Perih. I 98.23-24)

18a12

(G)M – H30

18a13-14 (In Perih. I 100.4-6)

18a17

HM

18a18-19 (In Perih. I 101.7-8)

18a23

GM – HO

18a23-25 (In Perih. I 102.17-20)

18a27

GM31

18a28-31 (In Perih. I 103.25-104.3)

18a33

GM

18a31-32 (In Perih. I 105.14-16)

18a33

H

18a34-36 (In Perih. I 108.14-17)

18a39

GM – H32

[In Perih. I 108.23: definite + 18a38-39 H] 18a39-b2 (In Perih. I 109.20-23)

18b5

GM – H

18b5-6 (In Perih. I 112.6-8)

18b7

GHM

18b11-13 (In Perih. I 114.5-7)

18b16

GHM

18b16-18 (In Perih. I 114.25-27)

18b20

GHM

18b20-21 (In Perih. I 116.1-2)

18b25

GHM

18b26-27 (In Perih. I 117.4-5)

18b36

GM – H33

18b36-37 (In Perih. I 118.11-13)

19a4

GM

[In Perih. I 119.7: mutabit + 18b39-19a4 H] 19a7-9 (In Perih. I 119.22-24)

19a22

GM – H

---------------------------------------------------------19a23-24 (In Perih. I 121.17-19)

19a28

H–M

19a26-29 (In Perih. I 122.19-23)

19a32

HM

19a32-35 (In Perih. I 124.8-13

19a39

M

19a35-36 (In Perih. I 124.29-30)

19a38

H

19a38-39 (In Perih. I 125.15-16)

19b2

HM

29 Cfr. above, ntt. 22, 25. Brackets indicate contracted interpolations or related disturbances. The complexities of these interpolations are such as to require separate study; the focus here is restricted to patterns of (dis)agreement in extension. 30 Cfr. above, nt. 25. 31 Cfr. above, nt. 21. 32 Critical apparatus and comments for this and the following lemma under III.B, below. 33 Critical apparatus and comments for this and the following lemma in Magee, Observations, pp. 20-23.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 37

27-Sep-21 8:05:54 PM

38

chapter ii

From which the following stemma emerges:

Ct φ

θ

χ

G H

O

M

The agreement of GHMO at 17b38-40 confirms φ, the lowest apparent stratum of interpolation. Whether the φ-compiler drew from C, t, or both is uncertain, although t appears the more likely of the two.34 φ comprises two sub-families, θ (GM) and χ (HO), which are most clearly in evidence at 17b20-21 and 18a18-19. O does not interpolate after 18a19, leaving the evidence for its descent from χ tenuous; indeed, O appears to have been purged of almost all connections with the φ-branch. The disappearance of G at 19a21 (In Perih. I 121.5) significantly alters the context for interpretation of HM thereafter: disagreements between the latter, a number of which lie below the level of their degrees of interpolation, and a lack of evidence in the earlier interpolations to indicate filiation along the lines of GH – M or HM – G, suggest that M continues to reflect θ at 19a23-24. Otherwise, evidence for the split between θ and H is strong (18a3-4, 3436, 39-b2, 26-27, 19a7-9) leaving φ-readings uncertain where ς-witnesses provide no assistance in adjudication between the two.

34 Cfr.

Minio-Paluello, [ed.] Arist. Lat., II, p. xlii, nt. 1.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 38

27-Sep-21 8:05:54 PM

CHAPTER III

EDITED PASSAGES The following passages have been reconstituted from the evidence provided by CDFGHKLMPV. O-variants have been relegated to the lower apparatus’ (lectiones singulares) in consideration of the manuscript’s severe contamination and ultimate irrelevance to the recovery of Boethius’ ipsis­ sima verba; for O inherits much confusion from its predecessors without shedding any independent light on the tradition. With one exception (L, 18a35-36), the singular readings of CDFGKLMPV have also been relegated to the lower apparatus’, along with five shared errors of minor significance (16b24; 65.19; 66.5; 109.6, 11); H-readings are reported in the middle apparatus’ (varia lectio) in consideration of the weight they carry vis-à-vis φ. The upper apparatus’ record differences of the lemmata and citations in c to the Greek (hereafter α), C, and t. Parentheses — (α), (C), (t) — signal points of disagreement in the evidence for each, for the details of which readers are referred to the critical editions of Weidemann (α) and Minio-Paluello (Ct). As noted earlier (II), the siglum ω signals readings of the archetype as derived either from the consensus of all ten witnesses or by recension (on which more below). On the non-interpolated side of the tradition (ς), the sigla μ and δ signal agreement between CKV and CV, respectively (cfr. II.A); and sigla for the interpolated branch (φ, θ) have been introduced above (II, II.C). [***] indicates a lacuna of approximately three characters, […] approximately three illegible characters, and /// an erasure of approximately three characters. Orthography has been normalized throughout. III.A. In Perih. I 65.9-66.28 3. 16b23 nec si hoc ipsum ‘est’ purum dixeris: 24 ipsum quidem nihil est, consignificat autem quandam compositionem, quam 25 sine compositis non est intellegere

65.13 ‘Est’ verbum iunctum cum alio verbo vel nomine 14 solet facere propositionem, ut ‘homo est’ vel ‘currere 15 est’, ipsum autem ‘est’ purum si dictum, inquit, fuerit 16 neque verum est neque falsum, a quo omnes paene 17 enuntiationes fiunt quae sunt simplices. ergo nec si 18 hoc ipsum ‘est’ purum dixeris esse aliquid aut 19 non esse significat, id est aut adfirmat aut negat, idcirco 20 quod ipsum ‘est’ simpliciter dictum nihil est, 21 non quod omnino nihil significet sed

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 39

27-Sep-21 8:05:54 PM

40

chapter iii

quod nihil neque 22 veritatis habeat neque falsitatis, id est non nihil est 23 ad significationem sed ad veritatis falsitatisque significationem, 24 de qua nunc tractabitur. cur autem nihil 25 veri falsique monstraret, ostendit. ‘est’ enim duobus 26 modis dicitur ut verum falsumque designet, aut cum 27 unam rem quamlibet propria compositione constituit, 28 ut cum dico ‘homo est’ ipsum ‘est’ cum homine 29 iunctum atque compositum esse hominem constituit 30 et fit exinde enuntiatio, aut rursus cum duae 66.1 res per ipsius verbi compositionem copulationemque 2 iunguntur, ut est ‘homo animal est’, ‘homo’ namque 3 et ‘animal’ copulantur atque iunguntur per id quod 4 dicitur ‘homo animal est’. ergo si omnis veri falsique 5 in enuntiatione significatio eius quod dicitur ‘est’ 6 in compositione est, cum aut sua compositione aliquid 7 esse constituit aut duas res copulat atque componit, 8 vis eius quae in enuntiationibus propositionibusque 9 monstratur, in veri falsique scilicet designatione, praeter 10 ipsas compositiones in quibus hoc solet efficere 11 nulla est. atque hoc est quod ait, consignificat 12 autem quandam compositionem, quam sine 13 compositis non est intellegere. cum enim dico 14 ‘est’, nihil significavi quid sit aut non sit, quod si 15 faciam enuntiationem, componam necesse est, ut ‘homo 16 est’ aut ut ‘homo animal est’. quare si omnis eius 17 vis in veri falsique enuntiatione ad compositionem refertur, 18 cum praeter compositionem dicitur quod in 19 compositione designare potest, id praeter compositionem 20 non designat, potest autem compositione facta verum 21 falsumque in enuntiatione monstrare. haec igitur 22 sine compositis in verbo ‘est’ nullus valet advertere. 23 recte igitur dictum est consignificare quandam 24 compositionem, quae sine compositis intellegi non 25 valeret. 4. 16b26 De oratione. Oratio autem est vox significativa, cuius 27 partium aliquid significativum est separatum, ut dictio, 28 non ut adfirmatio 16b23 hoc ipsum] αὐτὸ καθ’ ἑαυτὸ vel καθ’ αὑτὸ vel αὐτὸ vel quidem om. α | ‘est’] τὸ ὂν α(t) 25 non] οὐδὲν (α) 26 De oratione] (α) : quid sit oratio (t), cfr. infra ad loc. | autem om. (α) | significativa + κατὰ συνθήκην (α) 27 dictio + ἀλλ’ α 28 adfirmatio + ἢ ἀπόφασις (αCt), cfr. infra ad loc. Varia lectio: 16b24 autem CcDFKcLφ : om. LcPμ 65.15 purum si] Cc : si om. δ : pu. si [.] K : si pu. tr. G | inquit] Cc : -id CDP 20 dictum] om. μ : + est P 22 veritatis] CcKcMc : -itas CKMP | est2] sed aliquid praem. H : + sed aliquid est Hc 23 sed] δc : om. δ | ad2] Kc : a HK | significationem2] CH : scilicet nihil est CslHsl 24 tractabitur] F : -batur FcH 26 ut] GcV : aut GvidVc : aut cum DL | designet CcFGcHM : -at DGLPμ 30 fit] Vc : fix KV 66.4 si] GL : sic GcLc | omnis DFcGcHLMμc : -es FGPμ 8 eius quae] FcKcVc : -sque FKV | propositionibusque] Kc : -us quae KV : -us M : [****] pr- F 10 ipsas] δc : -a δ 11 consignificat] CcKc : non s- μ 14 significavi] Cc : -ativa P : -ativi δ : -at vi D : -avit Dc 16 ut FGHLcδ : om. DKLMP | omnis] GcKc : -es GKP | eius] HcKc : om. HK 17 vis in veri] μc : vis in vis in veri δ : vis inveniri KP | refertur … compositionem3 (19)] Cc : om. δ : compositione … praeter (19) om. P 16b26 de oratione (historia) oratio Cc(mg)DF(G) LmgP : [**********] oratio K : explicit de verbo incipit de oratione oratio H : oratio Mδ | est vox significativa] Lc : est om. P : vo. si. est tr. DL 28 ut] KcVc : om. DKV | adfirmatio] G : + vel negatio GcM Lectiones singulares: 16b23 nec] C : aristoteles praem. Cc | ‘est’] ?Pc : om. ?P | purum] Kc : par- ?K | dixeris] Gc :

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 40

27-Sep-21 8:05:54 PM



Edited Passages

41

om. G 24 quidem] Gc(vid) : -d est Gvid | est] Cc : om. C | consignificat] L : quia praem. LcO | autem om. O | quandam] Fc : quen- F 25 intellegere] C : + boetius Cc : + historia Gmg 65.13 alio] Lc : -iquo L 14 vel] Mc : om. M 15 est’1] esse K 17 enuntiationes] KcP : enti- K : -nis Pc(vid) | simplices] Vras | nec] Fc : om. F 18 aut] Fc : om. F 19 id est] idem ?K,M | aut1] Mc : om. M 20 est1 ip. tr. O 22 falsitas K | id est] idem M | est2] Dc : om. D 24 de] Lc(ras) | autem] enim G | nihil] Kc : om. K 25 monstraret] Cc : + et C | enim] autem O 27 una P 28 ut om. O | homine] Mc : -mo est M 66.1 per] Cc : om. C | verbi om. P | copulationemque] Cc : -m C 2 iunguntur] Kc : -git- K 3 et + homo O | adque M 5 enuntiatione] -ntitia[.]ne K : -ntiaone G | eius] ?LcOc : ei ?L,?O 6 compositione1,2] Kc : -nem K 7 est O | constituit] Fc(vid) : -uunt Fvid | co//pulat F 8 quae om. O 13 [*]um P 14 est + est O | quod] Gc : -id G | si om. P 16 quare si] Fc : -rsi F 19 compositione] Gc : po- G | designare] FLc : -ri Fc : om. L | compositionem] Gc : -io G 20 compositione] -em O : -sio- G 21 igitur] Gc : hi- Ginc 22 compositionis ?K 23 igitur] Mc : om. M 25 valeret] Hc : -let H 16b26 autem] Kc : om. K | cuius] Oc : cum Oinc 27/28 dictio, non] Fc : -onem Fvid 28 adfirmatio] C : + boetius Cc

The subject heading to c. 4 (16b26) has been discussed under II.A. As noted there, it may have been omitted by, and through correction reinstated in, an ancestor of CKV (μ); at the very least, V’s silence must be weighed against the correction to C and the lacuna in K. DFLP (hereafter ζ) and GH, by contrast, unambiguously confirm the presence of the title, De oratione, both below ς and in φ, thereby pointing directly to ω. Although there can be no guarantee that the subject heading was penned by Boethius himself, its antiquity is ensured by the evidence for t and Ammonius’ commentary (see above, p. 23, nt. 11). The framing formula for the first lemma (16b2325) has been discussed under II.B, to which the evidence here for 16b26-28 adds little: C (Aristoteles … Boetius) and G (historia) behave according to pattern. μ (Kδ): The evidence for μ and its corrector(s) is most conspicuous at 65.20 (omission of dictum), and 66.11 (non significat), with points of agreement between K and either C or V at 16b28 (omission of ut), 65.22 (veritas), 30 (fix), and 66.8 (eiusque, propositionibus quae) shedding more nuanced light: μ evidently contained a number of corrections which were inherited by Kδ; from those in δ CV and their correctors frequently selected independently of one another. Thus the ut at 16b28 would appear to have been supplied by a corrector of μ and so passed down to δ, whence it was adopted by C but initially resisted by KV. D’s shared omission there highlights a phenomenon evident also in many other passages, i.e., the circulation of the same errors and secondary readings beyond the immediate vicinity of μ, as at 16b24 (omission of autem), 65.26 (designat), and 66.4 (omnes). At 66.17 μ’s status before correction is obscured by the split between δ (vis in vis in veri) and K (vis inveniri), whose agreement with P may however be the tie-breaker. K’s agreement with H at 65.23 (a) and 66.16 (omission of eius) recalls its proximity (along with C) to θ at 16b18 (quod complectitur,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 41

27-Sep-21 8:05:54 PM

42

chapter iii

see under II, above); otherwise, its singular readings may safely be deemed irrelevant to the recovery of μ wherever δ joins with other witnesses (16b23, 26; 65.17, 22; 66.2, 5, 22). δ and its corrector come into view at 65.15 (omission of si, note the disturbances in GK), 23 (omission of sed), 66.10 (ipsa), 14 (significativi), and 17-19 (vis in vis in veri, omission of refertur … com­ positionem). ζ (DFLP): ζ is a provisional hypothesis which cannot be grounded on the independence of μ: DFLP clearly fall under ς but show no stable patterns of movement or coherence. Two points of agreement between DL stand out: 16b26 (vox significativa est, note the omission in P); 65.26 (aut cum). φ (Hθ): The absence of interpolated lemmata at this point in the commentary has a blurring effect on the visibility of φ, drawing attention instead to the readiness of GHM either to partner with ς-witnesses (e.g., 16b26; 65.22-23; 66.16) or to move independently (16b23; 65.14, 22, etc.). A secondary θ-reading is intimated at 16b28 (vel negatio), and H is enigmatic, with a number of intriguing symptoms such as the gloss shared with C at 65.23. III.B. In Perih. I 108.14-110.4 9. 18a34 Nam si omnis adfirmatio vel negatio vera vel falsa est, 35 et omne necesse est vel esse vel non esse; si hic quidem 36 dicat futurum aliquid ille vero non dicat hoc idem ipsum

108.18 Si quicquid in adfirmationibus negationibusque 19 proponitur verum vel falsum est definite, sequitur ut 20 quod illae negationes adfirmationesque significant aut 21 evenire aut non evenire necesse sit. hoc est enim 22 quod dicit, nam si omnis adfirmatio vel negatio 23 vera vel falsa est — definite, idcirco enim addidit 24 omnis. in his enim quae sunt futura vel contingentia 25 non esse adfirmationes et negationes veras 26 vel falsas definite probare contendit, nam si unus 27 dicat aliquid futurum alius neget, ut utraque eveniant 28 fieri non potest. quis enim dixerit, dicente alio 29 ‘Socrates cenaturus est’ alioque ‘Socrates cenaturus 109.1 non est’, utrasque veras in uno eodemque 2 contingere? hoc igitur fieri non potest, unus ergo 3 eorum verum dicturus est alius mentietur. ponatur 4 enim definite una earum vera esse vel falsa; si ergo 5 omnis adfirmatio et negatio vera vel falsa est definite 6 et uterque negans adfirmansque veri in contradictionibus 7 esse non possunt, necesse est unum verum dicere 8 alterum falsum et unum definite verum falsum alterum 9 definite. hoc si est in omnibus adfirmationibus 10 atque negationibus ut una definite falsa sit altera 11 vera definite, quicquid vera dicit eventurum necesse 12 est evenire, quicquid non eventurum non evenire necesse 13 est. hoc est enim quod ait, manifestum est 14 quoniam necesse est verum dicere alterum 15 ipsorum, neque enim potest fieri ut adfirmatio negatioque 16 consentiant in talibus, id est in contradictionis 17 positionibus. necesse autem esse evenire quaecumque 18 vera definite adfirmatio

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 42

27-Sep-21 8:05:54 PM

Edited Passages



43

loquitur et non evenire 19 quod profert definite vera negatio, sic probatur: 9. 18a39 nam si verum est dicere quoniam album vel b1 non album est, necesse est esse album vel non album, et si 2 est album vel non album, verum est adfirmare vel negare

109.24 Quod de futuro possit esse iudicium a praesentibus, 25 trahit exemplum. ait enim hanc esse rerum 26 consequentiam, ut rem subsistentem propositionis veritas 27 consequatur, veritatem propositionis rei de qua 28 loquitur propositio essentia comitetur. nam si hic 110.1 lapis vel quodlibet aliud album est, verum est de eo 2 dicere quoniam album est; et hoc convertitur, nam 3 si verum est dicere de eo quoniam album est, sine 4 dubio album est et album esse necesse est. 18a34 vel1] καὶ (α), cfr. 108.25, 109.5 35 omnia (C) | vel1] om. (αCt), cfr. infra ad loc. | si] ὥστε εἰ vel εἰ γὰρ vel εἰ δὴ vel εἰ δὲ α 109.13-16 manifestum … talibus] 18a36-39 18a39 vel + ὅτι (α) b1 esse + ἢ (α) | non2 + esse (C) 2 album1] λευκὸν Weidemann | est2] ἦν α, cfr. infra ad loc. | adfirmare] ἢ praem. (αCt), cfr. infra ad loc. Varia lectio: 18a34 vel negatio om. H 35 vel1 … futurum (36)] Cc : om. δ | vel1] Hc : om. H | si … ipsum (36)] Lc : om. L | si Meiser, cfr. 108.26 et supra ad loc. 36 ipsum + manifestum est quoniam necesse est verum dicere alterum ipsorum si omnis adfirmatio vel negatio vera vel falsa est utraque enim non erunt simul in talibus (18a36-39) θ : + media sententia est hic quae perficitur ex hoc quod paulo post infertur manifestum (-fie- ac) est quoniam necesse est verum dicere alterum ipsorum (18a36-37) H, cfr. ad loc. 23 108.18 quicquid] Gc : quid θ : quod Mc 21 est enim] Cc : enim om. δ : en. est tr. F 23 definite + utraque enim non erunt simul in talibus (18a38-39) H, cfr. ad loc. 18a36 24 omnis] CcDcFcLc : -es CDFL 25 et] vel in eadem lect. H 25/26 veras vel falsas] δc : -ra vel -sa δ 28 alio DLMPδ : al. quod Kc : al. /// quoniam F : al. quo K : aliquo GHLc 109.2 fieri / ergo] Cc : om. δ 3 ponatur] GcLc : -t L : -etur θ 4 unam GKLV | earum Meiser : eor- ω | vera CcDPφ : -am FLμ | falsa CDHKMP : -sam FG,?KcLV 5 est] CcKc : om. μ 8 definite] Cc : -tum Dδ 9 si est] DcVc : est si tr. D,?V 11/12 necesse est evenire Fc,?Kθ : ne. esset ev. D,?F,?KLP : ne. est et ev. H : non ev. ne. esset (est Cpc) evenire δ 12 non1] Cc : om. δ 15 fi. po. tr. θ 16 id est] CcKc : om. μ | contradictionis] FcVc : -ibus FV 17 positionibus CDFKLVc : propos- FcHLcPV : oppos- θ | autem esse] HcL : est es. P : est praem. Lc : au. (est inc) es. est H : esse om. M 18a39 dicere … verum est (b2)] Mc : om. M | vel … est1 (b1) CcDFLθ : est om. δ : vel al. non est tr. K : est vel non est al. H : non est albium est P b1 al.2 es. tr. H | album3] esse H 2 album2] est H | est2] erat H | adfirmare] Cc : -matio δ : vel praem. H | negare + et si non est mentitur et si mentitur non est quare necesse est aut adfirmationem aut negationem (+ de futuro H, cfr. 109.24) veram esse (+ vel falsam H, 18b1-5) φ 109.25 rerum] CKc : ver- K : veram Cc 110.1 vel] aut θ 3 de eo di. tr. H Lectiones singulares: 18a34 omnis] Lc : -es L | adfirmatio vel] Pc : -ione P | est + definite O 35 si] Dc : sic D 36 ipsus θ P 108.19 proponitur] Lc : pon- L | vel] Lc : om. L | definitae M 20 significant] Vc : -at V 21 necesse sit] Fc : -ssit F : ne. est P | enim om. O 22 negatio] Cc : nat- C 23 definitae F | addidit] Kc : addit P : added- K 24 omnis + ut ostenderet definitam G | contingentia] Mc : cong- M 25/26 adfirmationem … negationem veram vel falsam O 26 definite] Kc : -t K : -tae P | unus] Oc : unius O 27 al.1 di. tr. O | futurum] Vc : fur- V | aliusque G | ut] Kc : aut K : om. P | utraque] KVc : -roq- V : -quae P : tu Ksl | eveniant] Gc : -ientia G 29 alioque] F : -quod M : -iquisque Fc | Socrates[******]1 L | cenaturus2] Mc : caen- L : -tus M 109.1 non] Pc : om. P | eodemque] Lc :

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 43

27-Sep-21 8:05:55 PM

44

chapter iii

quid- Lvid : -quae F : + tempore G 2 hoc/// L 3 eorum] Mc : eo praem. ?M | est] Mc : om. M | alius] Dc : -iis G : albus D 4 enim] Mc : est praem. ?M | definite om. P | unam … veram … falsam O 5 vera] Lc : om. L | de. est tr. O 6 uterque] Kc : utraeque KO | in] Cc(vid) : an Cvid 8 falsum1] Pc : facere praem. P | alterum2] Kc : -ra K 9 hominibus P 10 definitae F | falsa … definite (11)] Cc : om. C 11 vera1] -ae D : om. O | definite] CcKc(vid) : -tae C,?K | quicquid … evenire1 (12) om. O | vera2] Dc : -aciter D 12 eventurus O 14 verum] Gc : -rbum P : om. G 15 ut] Vc : om. V | negatioque] KcVc(ras) : -o quae K 16 consentiant] Gc : -at G | contradic[*****]tionis L | in2] V : om. Vc 17 quaecumque] -m O : -mquae P 18 definite] Fc : -ta F 19 quod] Lc : -o L | differt O 18a39 quod albium P b1 est2] et ?P 2 est1] Lc : om. L | album1] Mc : ab- M | vel1] Mc : om. M | est2] Lc : om. L 109.24 possit] Kc : -set K | iudicium] Oc : -o O 25 trahit] -xit M : -ait P 26 subsistentem] KcOc : -ntium ?K : -nti O 27 propositionis] Lc : -sio- L : + veritas consequatur veritatem propositionis P 28 comitetur] Mc : conm- M : -itt- L | hil P 110.1 quodlibet] Kc : -ol- K | est1] Vc : verum praem. V | de eo ve. est tr. M 2 convertatur H 3 si] Kc : om. K 4 dubio] Dc : -i D : dubi/o L

ω (ςφ): Meiser rightly identified a false reading of the archetype at 109.4 (eorum); the error is probably an echo of (unus) eorum in the previous line. Meiser also supplied nam before si at 18a35 on strength of the paraphrase at 108.26 (cfr. the evidence for εἰ γὰρ, ap. Weidemann ad loc.). But the supplement is unnecessary, especially given the absence of any support from Ct, and the paraphrase more plausibly explained as Boethius’ effort to elicit Aristotle’s meaning from what appears to have been a uniquely elliptical Greek exemplar. μ (Kδ): μ has two patent omissions at 109.5 (est) and 16 (id est). Misreadings of rerum by CcK (veram, verum) at 109.25 reflect ambiguous letter formations (r, u) and, evidently, a correction or clarificatory note in μ. δ is most visible at 18a35-36 (omission of vel … futurum), 39 (omission of est), b2 (adfirmatio), 108.21 (omission of enim), 25-26 (vera, falsa), 109.2 (omissions of fieri, ergo), and 11-12 (non evenire necesse esset evenire, omission of non). ζ (DFLP): The theme here is again one of instability, as illustrated by the following patterns of agreement: CDFL (omnes), 108.24; GFLμ (unam, ve­ ram, falsam), 109.4; D,?F,?KLP (esset), 109.12; FcHLcPV (propositionibus), 109.17. P may well be the witness to watch most closely in this context. φ (Hθ): To 18a36 θ adds manifestum … talibus (18a36-39), overlooking the citation and paraphrase at 109.13-16. H splinters the interpolation, with the transposition of utraque … talibus (18a38-39) to 108.23 and a highly revealing editorial observation: media sententia est hic quae perfi­ citur ex hoc quod paulo post infertur, ‘manifestum’ ([etc.], “there is here an intervening passage, which may be pieced together from the citation just below, manifestum” [etc.], 109.13). Whether this is the voice of H or a precursor is uncertain, but H’s independence of θ is further attested by the omissions at 18a34-35 (vel negatio, vel) and the variants at 18b1-2. Most no-

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 44

27-Sep-21 8:05:55 PM



Edited Passages

45

table is erat vel at 18b2, where no witness other than H evinces awareness of the Greek imperfect (ἦν) but vel (ἢ) is firmly attested by Ct. Was there a medieval corrector with fresh access to the Greek or did erat vel descend from a collation note in φ or, indeed, ω? Either way, it is impossible not to discern the hand of a thinking, if misguided, editor here (18b1-2): necesse est album esse vel non esse, et si est album vel non est, verum erat vel adfir­ mare vel negare. To 18b2 φ adds et si … veram esse / vel falsam (18b1-4/5) to bridge the gap to 18b5 (In Perih. I 112.6); θ lacks vel falsam, which once again shows H (and one t-witness) in some form of contact with the Greek — and inadvertently dipping into the commentary (de futuro). Finally, in the commentary proper θ reveals its independence of H (or a predecessor) at 108.18 (quid), 109.3 (ponetur), 15 (fieri potest), 17 (oppositionibus), and 110.1 (aut). III.C. Conclusion In ω, the titles and colophons for each of the two books of the commentary were almost certainly written in Rustic Capitals1; the subject headings to cc. 2-4 were written in either Uncials or Rustic Capitals,2 as the framing formulae for lemmata and citations evidently were as well. For the lemmata and citations themselves, CKV (μ) have minuscules with pronounced traces of Uncials and rare indications of Rustic Capitals, while DFLP (ζ) vacillate between minuscules and Rustic Capitals with occasional traces of Uncials, and GHMO (φ) adhere strictly to minuscules; hence ς evidently had either Uncials or Rustic Capitals, as ω presumably did as well. A similar pattern is in evidence at In Perih. I 54.3-4, where FKV abbreviate the oblique Latin cases as follows: GEN, DAT, ACC (F) or AC (KV), VOC, ABL. In KV, GEN and DAT are written in Uncials that taper off into minuscules, whereas F adheres to a blend of Uncials and Rustic Capitals. CDLP exhibit minor deviations in abbreviation, employing minuscules for all but the letters G and D: CD have both letters in Uncials, whereas LP shift to Rustic Capitals for G. The φ-manuscripts have only minuscules, abbreviating quite independently of both one another and the ς-manuscripts. What evidently follows from this is that ς had Uncials, as presumably from ω, for genitivum and dativum but drifted into minuscules thereafter; it was followed by KV — i.e., by μ (which C was prone to clean up) — and normalized or “corrected” by F. CDLP adopted minuscules while preserving a 1 Thus CDGKLMO, with Uncial tendencies; Uncials with Rustic tendencies in P; UncialRustic hybrid in V; hybridized Square Capitals in FH. 2 Uncials in C (cc. 2, 3), FH (hybrid), KP (c. 2), V (hybrid); Rustics in C (c. 4, margin), D (hybrid), GLMOP (cc. 3, 4).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 45

27-Sep-21 8:05:55 PM

46

chapter iii

faint vestige of ς. For words and phrases written in Greek,3 the script used was certainly Uncial, as now in all but C, which at In Perih. I 32.9 displays its instinct for clarity and normalization with Caroline minuscules (perier­ menias). The logical diagrams at In Perih. I 86-87 are consistently reported by all but G, with its unique aesthetic flourishes, and P, with an unfilled lacuna for both; in ω they must have appeared very much as they do now in Meiser’s edition. There are other paratextual features which may also be confidently traced back to ω. Three passages, for example, point to ω’s use of itemization in Roman numerals as an exegetical aid. Thus at In Perih. I 71.1-5 DFGHKLPV express the series of ordinal numbers as prima, II, III, IIII, V,4 whereas CMO write all five out (prima, secunda, tertia, quarta, quinta). The uniform shift from prima to II, III, IIII, V cannot be coincidental but almost certainly points back to ω; for CMO’s version is in effect a lectio facilior probably devised independently by each. Similarly, at In Perih. I 79.23-28 CDF5KLPV (ς) interpolate ordinals expressed as Roman numerals, falsa negatio II et quod, falsa adfirmatio III et quod, adfirmatio vera IIII et quod, all of which are written out by C (secunda, tertia, quarta). Of the φ-witnesses, MO lack the ordinals altogether, while G has them above the line, and H has III and IIII in the text but II above the line. Most striking here is the complete absence of I (prima) as a primary reading: it appears only above the line in GK, which differ over its placement, and in the margin of V, with no signe de renvoie — three independent attempts to remedy what can only have been perceived as a patent omission. Did ς intercalate the three ordinals or did φ or a precursor remove them from ω? The latter seems the more probable hypothesis. Finally, at In Perih. I 81.12-13 CDFKLPV (ς) intercalate three cardinals expressed as Roman numerals: I quod si, II aut subiectum III aut praedicatum. They appear above the line in G but are absent from HMO. Here too it is difficult not to suspect ω as the point of origin. Such, then, are some of the glimpses offered by our manuscripts into the layout and text of ω, which may provisionally be traced back to Vivarium, c. 560-80. V breaks the silence of two centuries in providing a termi­ nus ante quem of c. 785 for the lower strata of editorial work and contamination discernible now in the extant witnesses. The evidence stimulates two complementary perspectives. On the one hand, corrections and glossing in CKV reveal dynamic processes of change which are datable to the 3 In

Perih. I 32.9, 62.19. omitted by GL before correction; III omitted by P and corrected from (?) vel in F. 5 Pace Meiser ad loc. 79.23. 4 V

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 46

27-Sep-21 8:05:55 PM



Edited Passages

47

early 9th c. but in certain cases reflect earlier phases of editorial activity in δ or (and) μ, as for example in the case of the (C)t-readings imported by early correctors at 16b18 and 17b32-33 (above, II). Where a wider array of manuscripts and corrections is in play, as with the confusion over the lemma, 16b6-7, and its associated subject heading (above, II.A), we may surmise editorial activity that occurred farther back in time. A more static impression is projected, on the other hand, by editorial changes that became fossilized in the text proper, as with the double readings in GHM (φ) at 17b32-33 (above, II). Passages in which V is a case in point furnish some of our deepest insights into the story of Aristotle’s emergence in the medieval West, as in the case of an embedded dittography at In Perih. I 49.23-26: designant (-at pc?) ergo id quod dicimur ferus cum alia parte nominis quae est equi unum consignificat equiferus separatum autem nihil extra designat ergo id quod dicimus ferus cum alia parte nominis quae est equi unum consignificat equiferus separatum autem nihil extra designificat

CK correct dicimur and avoid the dittography, but K has designant (23), and C designificat (26), before correction; from which it follows that both the errors and their corrections reach back beyond V to μ.6 A similar phenomenon has been noted in connection with the fractured lemma, 17a2-5, in L (above, II.C). Lorenzo Minio-Paluello astutely remarked upon the contaminated state of the oldest t-witnesses, the “Gallic” branch of the tradition whose initial intermingling with c he traced to Corbie, c. 820, and he rightly noted the high levels of contamination in the interpolated or “Germanic” branch of the tradition, especially as C gained in circulation from the 10th c. on.7 That the contents of the “Alcuinian Corpus” — more precisely, its inclusion of the Ps.-Augustinian and (?) Apuleian treatises — bespeak an unavailability of the Aristotelian Categories and De interpretatione at the time of compilation cannot be confirmed, however, unless it is demonstrated that δ and μ bear no traces of the latter — a highly improbable proposition. LO may safely be eliminated from consideration on grounds of the severity of their contamination, and the remaining nine witnesses organized as in the following stemma:

6 Cfr. Magee, Observations, pp. 16-17. 7 Minio-Paluello, [ed.] Arist. Lat. II, pp.

xli-xliii, xlviii-l, lii-liii.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 47

27-Sep-21 8:05:55 PM

48

chapter iii

ω

t(?C) φ

ς μ δ

ζ

θ V

FP D

C

K G M H

Clarity concerning the readings of μ and φ may be achieved with considerable confidence on the basis of the following patterns of agreement: (μ) Kδ, CK, KV; (φ) Hθ, GH, HM. ς, by contrast, is obscured by a lack of consistent patterns of filiation to confirm ζ. As to the contamination, one of the aims of this study has been to show that, although recension in its strictest sense is indeed impossible, there nevertheless are clear and consistent patterns of filiation which at many points hold the promise of leading back to ω via ς and φ. The remaining stages of investigation will be to ascertain any further patterns that may be indicated by the sixteen other extant manuscripts, most of which date to the 10th and 11th cc., and to probe the evidence for the second book of Boethius’ commentary.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 48

27-Sep-21 8:05:55 PM

CHAPTER IV

ARBEO, LEIDRAT, ALCUIN, AND THE CODEX PAGESIANUS The Vatican Library’s Pagès codex (CLA IV,417) is a manuscript of extraordinary importance for scholars of medieval philosophy, and there is already an authoritative bibliography dedicated to it.1 However, despite scholars’ outstanding contributions, we can say confidently that this very old and venerable witness to the rediscovery of Aristotle in the Middle Ages should be considered known more than properly understood. There is no doubt that most of the studies on this topic have highlighted a number of its meaningful aspects. However, these studies have also been unintentionally responsible for the stagnation of research and for channeling any further investigation along pre-set lines on the grounds of evidence that is not quite reliable. In fact, most scholars have accepted and developed the ideas of Léopold Delisle: the manuscript, offered as a gift by the bishop Leidrat to the church of St. Stephen of Lyon — as corroborated by the autograph signature2 — was most likely written in that city during his episcopate (798-814 A.D.). The manuscript seemingly depends on an archetype belonging to Alcuin, as proved by a quotation on f. 28r of the verses with which Alcuin dedicated to Charlemagne the pseudo-Augustinian Categoriae also included in the manuscript.3 In the pages that follow, we will try to point out the problematic aspects of this evaluation, which, in our opinion, should be contested, despite the authority of the scholars who have shared it. In our opinion, the date of 1 The

main paleographical and codicological contribution(s) regarding the manuscript are: Delisle, Notice, pp. 831-842; Martin, Note, pp. 119-20 (dating the manuscript to the VIII century); id., Invéntaire méthodique, pp. 486-87 (dating the manuscript to the IX century); Bischoff, n° 417 in Codices Latini Antiquiores (= CLA), IV; id., Karl der Grosse, p. 205, n° 363; id., Die Hofbibliothek Karls des Grossen, pp. 42-62; Aristoteles latinus. Supplementa altera, n. 2163, pp. 147-149; Kristeller, Iter Italicum, 6, p. 193; Klibansky-Regen, Die Handschriften, pp. 152-153 and Plate; Ward, The codex Pagesianus, pp. 308-333; id., Codex Pagesianus, in Codices Boethiani, III, pp. 309-311; Radiciotti, Romania e Germania, pp. 121144; Keefe, A Catalogue, p. 335; Munk Olsen, L’étude des auteurs clasiques latins, 4.2, p. 9. 2 «Leidrat licet indignus tamen episcopus istum librum tradidi ad altare sancti Stephani» f. 1v. 3 Poetae latini aevi Carolini, Carm. 73, p. 295.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 49

27-Sep-21 8:05:55 PM

50

chapter iv

composition of the Pagesianus should be moved to a time that precedes the ordination of Leidrat as bishop. The copyists’ handwriting leads us to draw this conclusion about the date of composition, since it does not show any resemblance to graphic experiments taking place at the end of the eighth century and at the beginning of the ninth. Rather, the handwriting seems to have much more in common with the eclectic and disorderly world of early medieval handwriting, typical of the second half of the eighth century. To this should be added a new element never before taken into account: the presence of a note of ownership on the flyleaf that appears to be an autograph and that is written by a hand typical of the eighth century. This hand appears in the margins of the manuscripts belonging to Leidrat (CLA 1269). The note of ownership underscores the fact that that at least the first booklet had belonged to Arbeo, bishop of Freising and teacher of Leidrat († 784), whose name is recalled at least two more times and by two more notes of ownership, written by contemporary hands on the front and back sides of the same leaf. Therefore, if the handwriting and the notes of ownership attest a relationship with Freising, then we are, at least for the first booklet, situated chronologically during the episcopate of Arbeo rather than the episcopate of Leidrat in Lyon. Beside Arbeo’s name, found at the beginning of the first work included in the codex, we find a note of ownership by Leidrat at the end of the texts included in the manuscript. Since each of these works has an independent fascicle numbering, later corrected in order to make it progressive and continuous, it is natural to interpret Leidrat’s notes of ownership as varying ex libris, each added to each single booklet independently from the others. These booklets, however, were written by the same scribes who wrote the first booklet, but with different inks, as happens when transcriptions are made years apart from each other. Only some pages were written by other copyists with handwriting very similar to that used by the first; this handwriting is reminiscent of the one found in contemporary witnesses typical of Bavaria. What we have briefly recalled means, therefore, that the Pagesianus was written by Leidrat’s trustworthy copyists, who followed him from the very beginning of his career, and was not brought to completion through uninterrupted work, but rather over the course of some years. The Pagesianus, therefore, is a “fictitious” codex: an artificial collection, made of different parts but transcribed by the same copyists at different dates between 784 and 798. This collection brings together uncommon works of logic, first copied separately and only later assembled into a single volume because of their undeniably deep and coherent relationship.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 50

27-Sep-21 8:05:55 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



51

I. Status quaestionis The starting point of the bibliography is an article published in 1897 by Léopold Delisle,4 who was contacted by the Marist Fathers of Lyon, the owners of the manuscript at that time. The illustrious French scholar displayed great expertise, worthy of his reputation and professional acumen, by carefully analyzing the manuscript, identifying the works and excerpts contained in it and endorsing the authenticity of the signature on f. 1v of Leidrat, bishop of Lyon.5 Delisle established a relationship between this signature and similar ones found in manuscripts that belonged to Leidrat and were donated to the church of St. Stephan of Lyon after Leidrat resigned; the French scholar later published the text. Delisle emphasized the peculiarity of the miscellaneous works contained in the manuscript, which was subdivided in two sections. The first section (ff. 2r-106r) includes old works on logic and reflects the pioneering interest in philosophy of the Carolingian cultural elite and of Alcuin in particular. The manuscript contains some abstracts of his De dialectica and the verses dedicated to Charlemagne. These verses were placed before some extracts of the translation of Aristotle’s Categories once wrongly attributed to Augustine.6 The second section (ff. 106v-114v) collects extracts from theological works and was conceived in the course of Leidrat’s polemics against the Adoptionist heresy and Felix of Urgell, related to his mission in Spain in 797-98 and 799. With the mentality typical of a positivist, Delisle applied the empirical results of his research to a general interpretation of the meaning of the manuscript in such a way that it seemed to emerge from the facts themselves and, as a result, would seem to be the most rational one. By applying the principle hoc post hoc ergo propter hoc, the French scholar painted a picture that looked logical and natural, but that, as a matter of fact, was founded more on appearance than on reality. Since the manuscript contained a section linked to the interests of Alcuin’s circle and another section linked to the missions in Spain of 797-98 and 799, and since it was donated to the church of St. Stephen of Lyon before 814, it seemed obvious to conclude that it was some 4 Delisle,

Notice, pp. 831-842. authoritative study on this topic is the following: Holtz, Leidrat, évêque de Lyon, pp. 315-334. A complete and updated study on Leidrat is still needed. Therefore, we simply refer the reader to Mordek, Leidrat, Bischof von Lyon, p. 1855 (Leidrat’s death is to be moved to a later time, since Claudius of Turin refers to him as still alive in 821: cfr. P Boulhol, Claude de Turin, pp. 34 and 55). See also the infomation collected in Gallia Christiana, Mètropole de Lyon et Vienne, pp. 95-101. 6 Edited in Aristoteles latinus, I, 1-5: Categoriae, pp. 129-175. On this work and the entire rediscovery of the logical works in early medieval Europe, see D’Onofrio, Fons scientiae, pp. 8-35, 277-328; id., Vera Philosophia, pp. LXXII-LXXIV. 5 An

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 51

27-Sep-21 8:05:56 PM

52

chapter iv

sort of summary of Leidrat’s intellectual biography, although Delisle did not explain in an explicit way what kind of relationship there was between the first section and the second. Essentially, the manuscript represents both the first phase of Leidrat’s life at the court of Charlemagne and the second phase of his career as missus dominicus in Spain and as bishop of Lyon, until he retired to a monastery after abdicating his office. Delisle consequently established that the date of composition of the manuscript, a manuscript which in a certain way summed up Leidrat’s biography, can be placed reasonably during his episcopacy and that the manuscript was transcribed in Lyon at the bishop’s orders.7 In this reconstruction, doubtlessly tout se tient, and the Norman paleographer must have been satisfied to have solved the disquieting puzzles presented by the codex. Yet many shadowy corners were still hidden from the light. In any event, this first contribution, offered by the most authoritative expert on manuscripts of his time, became a real milestone among the studies dedicated to the manuscript. Later work, however, helped scholars to identify with further accuracy some of the works contained in the manuscript, correcting Delisle’s initial conclusions. The manuscript was quoted and recalled on more than one occasion. Some pages of this old relic were reproduced and annotated in the collection of facsimiles of the confession of faith edited by Burn,8 as well as in the Monumenta Paleographica edited by Beer.9 Tafel, in the course of his research on witnesses from the scriptorium of Lyon, studied both Leidrat and the manuscript at length.10 During the same period, another reconstruction of the writing school of Lyon in Late Antiquity and the Early Middle Ages, even more brilliant and captivating, was added to his research: the volume on the Codices lugdunenses antiquissimi by the young but already influential Elias Avery Lowe. Of course, this volume did not discuss Leidrat, who was chronologically too late for it, but it did provide the basis for a great historiographical myth which, viewed retrospectively, would include the alleged activity of a cultural patron who inspired the transcription of works attributed to Leidrat, a bishop called filius noster by Alcuin. Through this type of research, an image took shape of a social and cultural center that had remained substantially stable over time, despite looting and devastation, underlining the continuity between characteristics typical of Lyon in late antiquity and those of the Carolingian period. 7 Delisle,

Notice, p. 838. Facsimiles of the Creeds, pp. 18-20 and plates XV-XVII.
 9 Beer, Monumenta palaeographica vindobonensia, pp. 47-49. But also see the text following the pictures. 11 a p. 46.
 10 Tafel, The Lyons scriptorium, pp. 66-73; 4 (1925), pp. 40-70, especially: vol. 2, pp. 6768, 70-73; vol. 4, pp. 51-52, 56, 63-64. The Pagesianus is mentioned in vol. 4, p. 52. 8 Burn,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 52

27-Sep-21 8:05:56 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

53

Within this general reconstruction, Leidrat’s profile was sketched in a way that inhibited new interpretations. It is not by chance that it could not play the role of main character in the glorious research through which the young Bischoff established himself internationally, namely the analytic and systematic study of the writing schools in southern Germany, particularly the production of the scriptorium in Freising and of the manuscripts belonging to its cathedral. Some of these manuscripts were written during the early youth of the future bishop. In his broad and well-documented reconstruction, the German paleographer was in some way obligated to take into account only the books and documents that the young Leidrat — deacon in Freising and at the service of the bishop Arbeo (763-84) — could have written, annotated, and owned in Freising, and to separate them from those that Leidrat would have donated to the Church of Lyon during his maturity, when he was bishop, after a long collaboration with Alcuin and Charlemagne. As a result, Bischoff could not dwell on manuscripts like the one we are dealing with, which, by common consent, was considered a product of the scriptorium of Lyon, completed during the last stage of the German bishop’s life. In any event, however, the great picture drawn by the young scholar highlighted the importance of the Mitarbeit of the young Leidrat alongside the bishop Arbeo, who actually created the scriptorium of Freising. It also identified manuscripts and characteristics belonging to the first stage of this very important writing center, as well as an alleged autograph of the future bishop of Lyon in Munich, Bayerische Staatsbibliothek, clm 6305 (CLA 1269). The time was right for a recapitulation of all that had been observed by scholars about the manuscript owned by the Marist Fathers with an eye to producing a final summary. The natural place for such a recapitulation was the great enterprise of the Codices latini antiquiores, to which the young Bischoff started to contribute in 1933 thanks to his teacher, Paul Lehmann, who recommended Bischoff to Elias Avery Lowe, then at Oxford. The L ­ ithuanian paleographer, who was finding it very difficult to work on this vast enterprise, needed collaborators who were up to the task and found some unanticipated help in Ludwig Traube’s young heir, coming from the same school at the University of Munich in which Lowe himself studied and earned his doctorate in 1911. Bischoff collaborated with extreme zeal in the preparation of the two volumes destined for the Italian libraries, traveling restlessly and collecting, according to his habit, a profusion of notes which allowed him to write most of the entries published in the third and fourth volumes of the CLA. The fourth volume, containing Roman manuscripts apart from those in the Vatican Library, could only be published after World War II. Bischoff returned to his collaboration

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 53

27-Sep-21 8:05:56 PM

54

chapter iv

with Lowe and was able to contribute to the final edition of the fourth volume in 1947. His name is unjustly absent from the front cover and appears only in a sincere but sibylline acknowledgment in the Foreword. However, his position was more important than we may be led to believe, and after his death his role in the drafting of the CLA was highlighted in the official bibliography of his immense scholarly production. The card file of the manuscript owned by the Marist Fathers, comprehensive and rich in information, relied, like many others, on first-hand work done by Bischoff,11 but was finally approved by Lowe, as was the case, after all, for many other card files which were originally written by the German paleographer and later taken up by Lowe, who at times rearranged or readjusted them. The file rephrased Delisle’s conclusions, according to the general principles of the CLA, by accurately describing the codicological structure of the Pagesianus. The German paleographer differentiated within it the handwriting of at least two scribes, of whom some peculiarities were noted, such as the use of a great number of Insular abbreviations. At the end of the description, he reviewed the witnesses to Leidrat’s signature, some of which were not reported by Delisle. It all seemed logical and rational; and for that reason, everyone liked it. And everyone echoed the same conclusions every time there was an occasion or need to do so. In fact, the information in the CLA entry had been accepted without hesitation in authoritative works such as the Aristoteles latinus, the Iter Italicum, and the catalogue of Apuleius’s works by Klibansky and Rege.12 Bischoff himself, after all, relied on what he wrote and consecrated it officially, both in the article that he wrote for the great exhibition on Charlemagne in 1965 and in his reconstruction of Charlemagne’s library, which appeared in the same article, defining Leidrat’s miscellany as “the rediscovery of dialectic as a teaching tool under Charlemagne … by Alcuin” and adding that “Der Kodex stellt die älteste erhaltene Sammlung dialektischer Lehrbücher dar… so spiegelt er die Wiederentdeckung der Dialektik als Unterrichtsfach und macht das Vorhandensein eines ganzen Corpus dialektischer Schriften am Hof wahrscheinlich, die Leidrat zum Besten der von ihm in Lyon organisierten Schule abschreiben ließ.”13 Obviously, the most authoritative studies on the history of philosophy in Late Antiquity and the Middle Ages, as well as on the culture of the Carolingian period, were based on the conclusions that had been reached 11 CLA

417 (cit. at nt. 2). To this end, see Krämer, Bibliographie, p. 184. nt. 2. 13 Bischoff, Die Hofbibliothek, pp. 42-62 (cit. at nt. 2). 12 See

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 54

27-Sep-21 8:05:56 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

55

unanimously by such distinguished paleographers and, relying on them, these studies provided clever interpretations of the meaning of Leidrat’s miscellany. The data of the paleographical and codicological analysis were indirectly corroborated by a number of authoritative scholarly works on the history of philosophy,14 which in turn depended in their inception on the same information that they were supposed to corroborate. Through a very peculiar question begging process, the history of culture, justified by paleographical investigation, in turn justified that which was the origin of its justification. Although the scholars produced works of great quality and value, at least as far as the Codex Pagesianus is concerned, everything remained in a closed circle that ended where it began. In other words, everyone in different ways repeated what had originally been said without verifying it, namely, that the codex of the Marist Fathers was produced during the episcopate of Leidrat and depended on the court library, or at least on Alcuin’s favorite works. But the ghost of the repressed — moping about, as Freud would say — lay in ambush. Leidrat’s shadow, like Banco’s shadow, was ready to appear suddenly: and this disquieting aspect hidden inside a reconstruction familiar to all — “unheimlich,” to quote Freud once again — made all of those scholars feel uneasy who still dared to think instead of being delighted by the cliché and feeling protected by an unshakable authority. In fact, under the apparently solid surface of an edifice erected with so much confidence by so many prominent scholars, cracks and inconsistencies remained hidden, hastily dismissed but sooner or later destined to resurface. II. De minimis non curat praetor The first contradictions were in Delisle’s article. The French scholar added to the renowned signature supposed to be that of Leidrat two very similar signatures derived from two contemporary manuscripts,15 stating that “le rapprochement de ces trois notes montre que touts les trois ont été 14 Some of the classical studies on this topic are, for example: Van de Vyver, Les étapes du développement, pp. 743-765; id., Opuscula: the Latin Aristotle, Amsterdam 1972; Marenbon, Tradition of Logic, pp. 49-50; Aristotle’s Peri hermeneias in the Latin Middle Ages, p. xxi; Marenbon, From the Circle of Alcuin, pp. 42, 52-55; Dales, Alcuin, pp. 57-58. For a general and updated introduction to the Carolingian period, see Bullough, Alcuin and the Kingdom of Heaven, pp. 37-40; Fried, Karl der Große. 15 One follows a fragment of the manuscript at the end of VIII, the Paris, BnF, lat. 152, and the other appearing at the beginning of a contemporary manuscript, the Lyon, BM, 608 (which in Delisle’s time had the number 524). Both the manuscripts are missing from the CLA.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 55

27-Sep-21 8:05:56 PM

56

chapter iv

tracèes par la même main.”16 But, in all truth, the three notes, although related to the same text17 edited by Delisle, did not seem to derive from the same hand at all: as we will see, there is an evident difference between them, both in the shape of the letters and in their structure (Plate 16). This latent contradiction by Delisle has never been carefully taken into account. And yet scholars must have had some doubts in this respect. Otherwise, we could not explain some problematic remarks in the CLA concerning other manuscripts in which Leidrat’s hand seems to be found. These remarks cast a hint of suspicion on the reconstruction proposed by Delisle, through a legitimate free association of ideas. In fact, in the sixth volume of the CLA, published in 1953, there is an explicit refusal to attribute to Leidrat’s hand the signature under his name in the manuscript 599 of Lyon (CLA 780). This manuscript (even in description 417 of the fourth volume of the CLA published in 1947) was long considered one of the books that belonged to the bishop. Since there also exists another manuscript (Lyon, BM, 610) in which the inscription of Leidrat has been added by a later hand — as is evident from the handwriting dating from the late ninth century and, after all, as Delisle quickly pointed out18 — it is possible to think that there may exist some non-autographic signature of the bishop and that, as a result, the impression of extraneity among the different hands in the images published by Delisle may be justified. I will deal later on with this topic. For now, I will limit myself to highlighting the contradictions of the apparently solid edifice built around Pagès 1. The most conspicuous contradiction is the following: Delisle subdivides the miscellany of the codex into two sections and highlights the fact that the second section, collecting many formulations of the Apostles’ Creed and other fragments, was linked to Leidrat’s mission against the heretics in Spain between 797 and 799: “La présence du symbole de Nicèe, de saint Ambrose, de saint Grégoire le Grand, de saint Grégoire de Néocésarèe, de saint Jerôme, sur les dernières pages du manuscrit, s’explique assez facilement: Leidrade avait dû réunir ces textes quand il passa le Pyrénées pour aller en Espagne combattre le doctrine de Félix, évêque d’Urgel”21. By in16 Delisle,

Notice, p. 839. licet in[dignus ta]men episcopus istum librum [tradidi ad altare] sancti Stephani», f. 1r, Lyon, BM, 608; 
«Leidrat licet indignus tamen episcopus istum librum tradidi ad altare sancti Stephani», f. 25r, Paris, BnF, lat. 152; «Leidrat licet indignus tamen episcopus istum librum tradidi ad altare sancti Stephani», f. 1v, Biblioteca Apostolica Vaticana, Pagès 1.
 18 «Leidradus episcopus istum librum tradidit ad altare sancti Stephani. Qui furto exinde sublatus et inventus alia manu 
novo titulo insignitus est, quia prior, sicut evidenter agnosci potest, a furti auctore abrasus fuerat.» f. 1r: cfr. Delisle, Notices sur plusieurs anciens manuscrits, pp. 390-391. 17 «Leidrat

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 56

27-Sep-21 8:05:56 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



57

advertently applying the principle hoc post hoc ergo propter hoc, the French scholar devises an apparently sound criterion for the dating of the manuscript that is based on its content. And everyone, seduced by the seeming soundness of the argument, followed him down this path, even the most recent scholars, as ostentatiously subversive in their words as they are conformist in their deeds. And yet the reasoning is flawed in many respects. As we know from Burn’s pioneering research in 1909, corroborated by recent studies, it was rather common in the Carolingian period to gather various formulations of the Creed into Carolingian manuscripts, without regard to the controversy against the Adoptionists. It is not by accident that Burn himself expressed his skepticism concerning Delisle’s opinion when he wrote: “Delisle suggested that the collection of Creeds was prepared for Leidrat’s journey to Spain in 798, when he was combating the heresy of Adoptianism. But there were no phrases even in the Quicumque which directly combated this revival of Nestorianism. It seems more probable that the collection was the fruit of the general impulse given to historical research and theological studies by the influence of Charles the Great. We find similar collections in other manuscripts of the period at Leyden (Co­ dex Latinus XVII 67 F, saec. VIII-IX) and at Karslruhe (Codex Augensis XVIII, saec. IX in.). But it is of more importance to note that exactly the same collection of Creeds, in the same order, together with most of the abstracts which follow in Leidrat’s manuscript, actually form the introduction of the famous Golden Psalter of Vienna (Cod. 1861)19 which was written by the command of King Charles for Pope Hadrian.”20 To these considerations the following should be added: even if we wanted to link the miscellany of formulations of the Creed to the Adoptionist controversy, we would not be bound to date the manuscript to 798. In fact, already in 785, Pope Adrian explicitly criticized Spanish Adoptionism, and Felix of Urgel tried in vain to appeal to Charlemagne but was condemned at the Synod of Regensburg in 792. As is clear by now, the “battle of the Creeds,” to use Douglas Dales’ expression,21 far precedes Leidrat’s missions in Spain between 797 and 799 and, as a result, the miscellany of formulae fidei of the manuscript of the Marist Fathers could have been prepared before the period from 797 to 799. Apart from this, a further consideration leads us to reject the idea that the manuscript was written after 798: the texts of Leidrat’s miscellany were 19 This

is the famous Psalter of Dagulf, written before 795, about which see Bischoff, Paleografia latina, pp. 161-162 (with bibliography). 20 Burn, Facsimiles, p. 19.
 21 Dales, Alcuin, pp. 70-74.


www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 57

27-Sep-21 8:05:56 PM

58

chapter iv

employed in the Libri Carolini, dating from 790 to 795, well before his mission to Spain.22 But however desirable further research on this topic may be, at present there is no need to deal with this question and enter more deeply into the details of the intricate issues related to this work and the discussions taking place in Carolingian times concerning the Creed, the formulation of the Filioque,23 and its influence on the so-called Dicta Albini, whose oldest exemplar is precisely this manuscript, Pagès 1. For us, it is sufficient to emphasize the fact that Leidrat’s miscellany, used in the Libri carolini before 794, is not necessarily connected to his trip to Spain between 797 and 799. Of course, the future bishop of Lyon used the texts gathered in his codex during the controversies in Spain, and this, as we will see, can be proved thanks to a detail missed by scholars. Yet this does not mean that his miscellany was prepared only and exclusively in view of his preaching against heretics. Rather, it mirrors the cultural tensions and controversies of the last decades of the eighth century, attested both by the conflict between the Spanish heretics and the Carolingian court and by disagreements between Charlemagne and the Byzantine Empire revealed in the Libri carolini, which, of course, also have consequences for the controversies surrounding Spanish Adoptionism. To sum up, we must admit that it is impossible to date the Pagès codex after 798 on the ground of the collection of formulae fidei in its final section, since this collection originates from issues that surfaced throughout the period from 785 to the end of the eighth century. The change in the dating of the codex also implies a change in the evaluation of its meaning and origin. By dating the manuscript to the time of Leidrat as bishop of Lyon, Delisle, followed by everyone else, attributes the volume to a scriptorium of Lyon, and holds that it was copied in order to pass on to the cultural elite of the city texts and knowledge that had previously been restricted to the Carolingian court. But if it is no longer certain that the Pagesianus was written after 798, neither is it certain that it was copied in Lyon with the purpose of introducing the culture of the Carolingian court to the city. There was one way in which this doubt could have been resolved once and for all: comparing the Codex Pagesianus with other codices belonging to the same context in Lyon. If points of contact between all of them could 22 Meyvaert, Medieval notions, pp. 78-87, in particular pp. 86-87. On the Opus Caroli see, in general, the introduction by Anne Freeman to the Opus Caroli Regis, pp. 3-92.
 23 Fully aware of simplifying a very rich bibliography, but also of the very limited space allowed to our study, we limit ourselves to mention of recent research, which also includes bibliographical references: Gemeinhardt, Die Filioque-Kontroverse.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 58

27-Sep-21 8:05:56 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

59

be found, then these doubts would be dispelled. But in their absence, how can we hold that Leidrat’s volume was really copied in Lyon? In truth, it is not by chance that such a comparison was never made. Identifying old products of the scriptorium of Lyon with certainty is very difficult, despite the efforts of many excellent scholars.24 The results of such efforts have been questioned more than once by authoritative voices who criticized the naiveté with which manuscripts were attributed to the scriptorium of Lyon merely because they had belonged to the cathedral of Lyon on the grounds of some sort of “Hegelian historicism” very similar to the one that inspired Delisle. In a very clear article,25 Philippe Laurer criticized the empirical approach explicitly formulated by Lowe, according to which: “Si un manuscrit très ancien est conservé dans un centre calligraphique qui a eu un développement antérieur, on est fondé à croire, à défaut de preuves contraires, qu’il est originaire de ce centre.” Laurer pointed out the weakness of such arguments. He noticed, among other things, that it is difficult to identify products of the scriptorium of Lyon, both because there is not any handwriting typical of this center and because we know that many manuscripts that belonged to the cathedral or to other institutions in the same city came from Italy, Spain, and other regions in France. Nonetheless, Laurer’s remarks remained a dead letter and were never able to prevent the most widespread opinions from being repeated, as was natural in studies concerning people like Florus of Lyon and Agobardus, who owned and annotated manuscripts once owned by Leidrat. But the contradictions hidden beneath the vail of apparent soundness were not eliminated; rather; they resurfaced in many studies. Let us mention, for example, the obvious contradiction in the entry about the Pa­ gesianus in the CLA: after a few lines, the entry classifies the handwriting of the manuscript as a true “Carolingian minuscule” and right after “Early Carolingian” handwriting, underscoring aspects of the handwriting not fully developed. The entry, moreover, is very cautious in the attribution (“presumably”) of the manuscript to Lyon, leaving room for future research. Later on, in 1965, Bischoff called the handwriting of the manuscript “frühkarolingisch,”26 correcting with a single expression the previous contradiction: this definition is part of the general ­evaluation which the author made, in the same year, of the book pro24 For a first introduction to this topic, see: Lowe, Codices Lugdunenses antiquissimi; Bischoff, Frühkarolingische Handschriften, pp. 306-314; Vezin, Manuscrits présentant des traces, pp. 157-171; Holtz, L’évêque, le diacre et le manuscrit, pp. 28-34. 25 Lauer, Observations sur le Scriptorium de Lyon, pp. 380-387. 26 B. Bischoff, Karl der Grosse: Werk, p. 205, n° 363 (cit. at nt. 2).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 59

27-Sep-21 8:05:56 PM

60

chapter iv

duction of Carolingian Europe in a study which has remained famous.27 However, we should emphasize that Bischoff’s positions are only a reflection of the bibliography available on the topic before 1965 and of nothing after this date: it is therefore rather naive — and it was, in fact, done — to appeal to the name of the German scholar after 50 years while deliberately disregarding or ignoring scholarship on Carolingian minuscule more recent than the 1960s. Bischoff returned to the Pagesianus once again in 1965, in the catalogue of a large exhibition on Charlemagne. Obviously, the circumstances of an exhibition were not favorable to an in-depth study of the Pagesianus, and the entry prepared for this event reflects, even in the vocabulary, the one already prepared for the CLA. I am not saying, of course, that Bischoff did not realize the importance of the codex: it is not an accident that he spoke about it in his essay on the Library of Charlemagne. Nevertheless, it is in no way fortuitous that he did not dedicate any further attention to the manuscript,28 apart from the occasional and uncommon curiosity of a probatio pennae in the upper margin of f. 109v, where a ninth- or tenth-century reader wrote the words of a Psalm upside-down.29 Bischoff’s indifference was certainly conditioned by his appreciation of Delisle and by his deference to Delisle’s authority. There was no reason, from his point of view, to question authoritative opinions on the Codex Pagesianus. But besides this, a second impediment, perhaps unconscious, held him back and represented another difficulty. The Roman manuscript, in fact, presented a problem whose solution was very difficult and could only have been solved through a very thorough study: the identification of Leidrat’s handwriting in the marginal notes of the manuscript. The presence of such marginalia was considered a matter of course, but their identification was not, just as it was not at all obvious that such annotations were similar to the initial signature. A problem of this kind concealed another: how should one evaluate the written and personal culture of a man like Leidrat? And what kind of link was there between Leidrat’s cultural and writing activity in Lyon, as witnessed by the manuscripts that he annotated and signed, and his activity 27 B. Bischoff, Panorama der Handschriftenüberlieferung aus der Zeit Karls des Großen, in

Karl der Grosse. Lebenswerk und Nachleben, 2, pp. 233-254 (cit. at nt. 3). 28 This is clearly shown by the Plate of the manuscripts of his bibliography prepared by Krämer, Bibliographie, in which the manuscriptx appears only on pp. 28 and 38 (apart from the mention made in the entry 417 of the CLA).
 29 Bischoff, Mittelalterliche Studien, 3, pp. 122, 157.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 60

27-Sep-21 8:05:56 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



61

during his youth in Fresing, as witnessed by the surviving manuscripts from the cathedral library? The noricus Leidrat, whom many remember only because he was bishop of Lyon and a disciple of Alcuin, was, as a matter of fact, educated in Freising and was a personality who could not be ignored in Munich, or worse, dismissed as a clergyman working for Charlemagne. Leidrat began his career as deacon of the church of Freising and notary of the great bishop Arbeo, on behalf of whom he prepared documents, one of which was dated 779 and preserved in a later transcription.30 Bischoff referred explicitly to this document and to this activity in his outstanding research on writing centers in Bavaria, also mentioning another document written by the same Leidrat in 782,31 this time, however, not while working for the bishop, but rather for the Tassilone who sent Arbeo into exile in the same year because the bishop of Freising sided with Charlemagne. It is obvious from this action that Leidrat did not follow his bishop into exile but remained in the service of the Church of Freising and Atto, who served as bishop until 784, officially replacing Arbeo after his death. Bischoff was well aware of the situation, as were all of the scholars who dealt with the history of Munich and of Bavaria. And, like all of them, the renowned German paleographer desired eagerly to recover the relics of one of the founding fathers of the library in Fresing and of Bavarian culture. It is no coincidence that, by reconstructing Arbeo’s scriptorium and drafting a list of the volumes that survived in the library of Munich, ­Bischoff brought together three manuscripts that dated to Arbeo’s last years or Atto’s first years in which it was possible to think of Leidrat’s “Mitarbeit”: the manuscripts Munich, Bayerische Staatsbibliothek, clm 6303, 6305, and 6309 (CLA 1268, 1269, 1272). 6305 in particular was suspected to show “die Mitwirkung Leidrats.”32 It raised not a few expectations in the community of scholars in Munich, which counted very much on the “divining” ability of Lehmann’s pupil in order to provide a solution to the question. Bischoff did not disappoint their expectations. After stating in the entry of the clm 6393 on p. 83 that “[das] clm 6305 … wir mit Leidrats Namen versehen dürfen,” he stated that, in 6305, written “in der letzte Jahre Arbeos oder in Anfang die Regierungszeit Attos,” he believed that he had found the presence of Leidrat’s handwriting on lines 1-19 of f. 130r of clm 6305: “Wie die Schrift in clm 6305 in Erscheinung tritt, habe ich kein Be-

30 Bitterauf,

Die Traditionen des Hochstifts Freising, pp. 113-114, n° 95. pp. 122-123, n° 106. 32 Bischoff, Die südostdeutschen Schreibschulen, p. 64. 31 Ibid.,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 61

27-Sep-21 8:05:56 PM

62

chapter iv

denken, in ihr die Hand Leidrats … zu sehen.” This handwriting had also been found on ff. 138v-141v of clm 6393 (CLA 1280).33 Consequently, it might be expected that everyone, and Bischoff above all, would be content with the result achieved. In a certain way, that is what happened.34 And yet a gnawing doubt tore at the German paleographer and incited him to write veiled retractationes. The voice of his conscience whispered to him not to be so sure of the identification of Leidrat’s handwriting, the voice of his conscience and his more or less accurate memory of the manuscript of the Marist Fathers. Bischoff could not ignore something quite evident to a true paleographer: the manuscript Pagès 1 posed many problems. Undoubtedly, the hand that wrote the dedication was similar to the annotations reproduced by Delisle; but there were also annotations by another hand, present in the margins of clm 6305, which were very different from the writing of the dedication. Later on, I will dwell at length on this point: for now, I will only mention it and remind the reader that Bischoff, unsatisfied with his own conclusions, changed his mind a few years later without showing it, patently contradicting himself. Bischoff’s veiled retractatio was written a few years later while preparing the ninth volume of the CLA. In fact, in the entry of the clm 6305 was written: “the script of f. 130, lines 1-19 … recalls the hand of Bishop Leidrat…” Recalls here just means “reminds one of,” surely not “is to be identified as.” This time, Bischoff does not feel confident anymore to write: “ich habe kein Bedenken, in ihr die Hand Leidrats … zu sehen.” If we wish to tell the whole story, Bischoff was also brooding on something else. We have already recalled his refusal to acknowledge Leidrat’s autograph in the dedication of Lyon, BM, 599 (CLA 780). This refusal was a sign of a deeper unsettled issue: are all Leidrat’s autographs really genuine? Bischoff never openly advanced such doubt, but without ever asking himself such a question he never would have ventured to doubt the autograph of Lyon, BM, 599. There is more. In his entry in the CLA, the young Bischoff carefully listed all the Insular abbreviations and the different Insular symptoms of the two copyists of Pagès 1, who constituted a sort of background plot to the non-Insular script. A similar coexistence of writing methods can be found in several centers at which the presence of Insular scribes is doc33 Ibid., pp. 83-85. The quotation is on p. 84. This opinion is located in the most recent catalogue of the manuscripts of Freising preserved in Munich: see Glauche, Katalog der lateinischen Handschriften, 1, p. 184. 34 It is not by chance that even in 1949 Siegmund, describing this manuscript, stated, quoting Bischoff: “Vielleicht hat sogar Leidrat daran mitgearbeitet” (Siegmund, Die Über­ lieferung, p. 136).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 62

27-Sep-21 8:05:56 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



63

umented, as, for example, Freising itself at the time of Arbeo, where an Insular copyist who signs his name “Peregrinus” transcribed exemplars and dedicated them to the bishop (CLA 1253, 1263, 1265). Likewise, we can observe the same phenomenon in other centers open to Insular influence, such as Saint Gall or Reichenau. But as far as Lyon is concerned, or at least the codices rightly or wrongly attributed to the scriptorium of Lyon, this eventuality has not been reported. How, then, is it possible to reconcile the belief echoed by all and shared by Bischoff, that Pagès 1 had originated in Lyon, with the presence of considerable influence from Insular scripts on his copyists? To pretend that this contradiction does not exist involves a number of absurdities. To begin with, even the most expert scholars asking serious questions about the codicological and paleographical aspects of the manuscripts apparently originating in Lyon are forced into all kinds of acrobatics in order to balance accounts which never add up. This is the case with Jean Vezin, author of a famous essay on La réalization materielle des man­ uscrits latins pendant le Moyen Âge.35 The French paleographer provided a study of the highest quality of Pagès 1, drawing attention to a codicological peculiarity common to a group of manuscripts from the Early Middle Ages: the custom of beginning the fascicle on the flesh side. Vezin reminds us that such a peculiarity is found in Insular manuscripts that influenced many scriptoria in which scribes coming from the islands were active, such as, for example, Freising, Lorsch, Reichenau and Salzburg. Since this way of presenting the fascicle is also adopted in Spain (as well as some centers in the Northern part of France), Vezin ended up drawing the conclusion that the Codex Pagesianus must have been influenced by Spanish habits, because, according to the traditional opinion, the codex was composed in Lyon where Spanish scribes were active. Now, there is no doubt that there were Spanish scribes in Lyon. It was no coincidence that Leidrat’s successor was Agobardus, who was originally from Spain, and it was no accident that Felix d’Urgel, who was Leidrat’s major opponent during his mission to Spain, was deported to Lyon and kept under surveillance. Yet the presence in Pagès 1 of a codicological peculiarity very common among the scriptoria influenced by Insular scribes and shared by Spanish ones — the fascicles beginning on the flesh side — should and indeed may have been connected with the presence in the same codex of another element linked to Insular scribes, this one of a paleographical nature, namely the considerable presence of Insular abbreviations not found among Spanish scribes.

35 Vezin,

La réalization materielle, pp. 15-51, in particular pp. 26-27.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 63

27-Sep-21 8:05:56 PM

64

chapter iv

III. Nova et vetera The “return of the repressed” did not have the result one might have anticipated. Instead of entering into a crisis when faced with the surfacing of blatant contradictions, many reacted by reaffirming without further discussion the traditional views with their lack of clarity, adopting a defense mechanism consisting of removal and denial. In 1993, Anthony Ward published an essay dedicated solely to the Codex Pagesianus, the first since Delisle’s essay ninety-five years prior.36 Theoretically, his article should have picked up the threads of the question and properly reformulated all of the issues in light of an updated bibliography. Practically, however, the scholar did not manage to wriggle out from under the weight of tradition but wrote an article that does not fully meet expectations, although being worthy of attention in many respects. Ward doubtlessly helped scholarship make some progress by adequately reconstructing the history of the manuscript and the collection of the Marist Fathers, as well as by finally correctly identifying all of the texts contained in the manuscript. Another step forward was made by situating, with the aid of an updated bibliography, the miscellany of Creeds in the Pagesianus within the context of the religious controversies that characterized the Carolingian court between the end of the eighth and the beginning of the ninth century. Nevertheless, Ward remained bound to Delisle’s opinions and, in fact, ended up justifying the latter’s auctoritas even when he dared to express some doubt, writing: “Even if we do not subscribe to Delisle’s oft-repeated hypothesis that a collection of Creeds as found in the Pagesianus was put together specifically for the use of [Leidrat’s] expedition [in Spain], the general interests of the epoch are still clear, and Delisle is right to point out that Leidrat was in general terms well-documented for this controversy, the armoury including writings of Alcuin.”37 In order to corroborate this idea the scholar adds, in a remarkable Pindaric flight, that the presence of Juvencus’ verses in the manuscript (not identified by Delisle) stems precisely from this anti-Adoptionist tension, since “Juvencus was also of Spanish origin, and in his more conciliatory moments we observe Alcuin in particular trying to argue against Adoptionist positions on the strength of authors of Spanish origin who might gain a readier hearing among their countrymen.”38 This statement seems to be justified neither on the grounds of what is known about the Adoptionist controversy nor 36 Ward,

The codex Pagesianus. p. 331. 38 Ibid., p. 332. 37 Ibid.,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 64

27-Sep-21 8:05:56 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

65

on the grounds of what is known about Juvencus’ fortune in the Middle Ages.39 Ward tries to explain the presence of Boethius in the collection by invoking religious motivations similar to those that inspired the collection of various formulations of the Creed. He writes, “Even the philosophical texts in the Codex Pagesianus form part of a web of thought that leads in and out of texts (formally absent, but not without their influence) such as Augustine’s De Trinitate. Thus they relate to faith and its articulation. A little reflection will also bring to light how strong are cross-links between the different categories of material contained in the codex. Boethius, for instance, is everywhere in the first and principal part. We can hardly overlook the fact that in the psalter-related material that largely includes it, we find passages from Cassiodorus, linked to the same Boethius by family kinship through the celebrated Christian Anici, by high admiration unstintingly acknowledged, by outlook and by a range of activities.”40 We cannot agree with such a judgement, apologetic and apt to reduce the importance of the rediscovery of Boethian logic, which was the object of authoritative studies in the wake of the tradition of the Aristoteles latinus. After all, the collection of the logical works in the Pagesianus has little to do with religion, since it is composed of pagan authors like Apuleius and commentaries and summaries of another pagan author, Aristotle. After Ward’s intervention followed, about fifteen years later, an article of a paleographical and codicological nature: following the example of the Marist scholar, Paolo Radiciotti published in 2008 an essay once again entirely dedicated to Pagès 1.41 The Roman scholar reiterated with determination everything that had already been said about the Pagesianus and added a series of more general observations, considering the manuscript an excellent witness to the birth of the Carolingian script. Radiciotti’s contribution has received severe criticism from authoritative paleographers, who openly dissociated themselves from the author’s misunderstandings and from his methodological approach, as well as from his theory about the origins of Carolingian script, which leads to more than one “worrying and scary”42 statement, and, in any case, “is not fully 39 On this topic, which has been dealt with many times by scholars, we refer, as a simple example, to the recent volume by Green, Latin Epics. 40 Ward, The codex Pagesianus, p. 333. 41 Radiciotti, Romania e Germania. 42 Supino Martini, Aspetti della cultura, pp. 213-250, in particular p. 224: “Sembra che Radiciotti cada nell’equivoco che una merovingica poco cancellerescamente atteggiata e con lettere spaziate possa definirsi protocarolina”; p. 244: “Radiciotti … conclude con la definizione di un altro inquietante quanto improbabile nesso tra merovingica e carolina”.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 65

27-Sep-21 8:05:56 PM

66

chapter iv

convincing.”43 Such severity may appear excessive, but it does not, in fact, differ from the negative evaluations that other prominent scholars made of other contributions by Radiciotti.44 Without delving too deeply into these evaluations and Radiciotti’s theories about the origins of Carolingian writing, I will limit myself to observing — remaining within the boundaries of my research — that the essay is, paleographically speaking, rather superficial, both because it does not take into account the issues raised by the literature — which at times the author, as is his custom,45 does not know at all46 — and because, despite his initial promises, he neglects the paleographical and codicological analysis of the manuscript in favor of general enquiries, rather fuzzy and unjustified, concerning the origins of Carolingian script. Is it not it strange, for example, that the author should state categorically that the manuscript is the “fruit of the copying activity of a single scribe,”47 rejecting without explanation Bischoff, who talked about two scribes, while the images that he published in the article were, without a doubt, images of the handwriting of two different hands, using different letters, different abbreviations, and placing words on the line in a way that is evidently different? And is it not strange that the author writes an entire essay on a manuscript in “Early Caroline,” when the scribes who transcribed this manuscript do not make use of “Early Caroline” at all? A similar contradiction should be obvious to the author himself, since he rightly states that “the most important critical element” in identifying the Caroline script is “the distinction between manuscripts written with ligatures and … those, on the contrary, where the ligature is a marginal presence.”48 But if this is true, how is it possible to consider the handwrit43 Pratesi-Cherubini, Paleografia latina, p. 365: “Non convince pienamente la recente tesi di Paolo Radiciotti per cui la nascita della carolina sarebbe stata funzionale ad abbreviare i tempi di educazione degli scribi.” 44 According to Giuseppe De Gregorio, Radiciotti’s evaluation of Paris BnF, gr. 54 is “metodologicamente insostenibile” De Gregorio, Tardo Medioevo, pp. 43-44, nt. 56). According to Luciana Cuppo Csaki, the dating of the Vat. lat. 3764 proposed by Radiciotti is “metodologicamente carente e non suffragata da adeguata documentazione paleografica — quindi francamente insostenibile” (Cuppo Csaki, Pontifices di Costantinopoli, pp. 359-372 in particular p. 365, nt. 101). 45 A bad habit rightly censured by scholars. See the criticism of this issue by AmmannatiStagni, Ancora sul manoscritto, pp. 399-424. 46 Let us think, for example, of the abscence of consultation of Burn’s work, which we have previously recalled, and of his critical observations of Delisle regarding the collection of different formulations of the Creed, which more recent studies on the same topic (also missing from Radiciotti’s notes) have widely confirmed. 47 Radiciotti, Romania, p. 132. 48 Ibid., p. 134 (cit. at nt. 36).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 66

27-Sep-21 8:05:57 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



67

ing of the copyists of the Pagesianus “original Caroline” or “proto-Caroline,” when theirs is a handwriting full of ligatures which are not at all “a marginal presence,” but, on the contrary, a continuous presence49; a handwriting whose appearance changes from one line to the next, a handwriting that at one time shows long vertical strokes and a small central body, and at other times short vertical strokes and an enlarged central body, a handwriting in which, unlike Caroline, no single paleographical and alphabetical ideal is represented and the letters themselves are shamelessly written in different ways50? The Caroline script originally presented only some alphabetic variation — such as the alternation between “a” with an open, curved stroke and “a” with a closed, curved stroke — not the paleographical mess and continuous fluctuation among the different solutions adopted by the scribes of the Pagesianus. On this issue, Armando Petrucci wrote: “The Carolingian minuscule was a minuscule round in its forms, simple and balanced in its pattern (based chiefly on a harmonious relationship between the body of the letter and the ascenders, between the horizontal extension of the loops and the vertical extension of the high ascenders), airy (with a clear separation of the letters rather than of the words among themselves), almost devoid of ligatures and abbreviations.”51 The most recent and authoritative handbook of Latin paleography, written by an authoritative scholar such as Alessandro Pratesi, and by one of his young and brilliant colleagues, Paolo Cherubini, is of the same opinion: “The first characteristic of a Caroline script which immediately catches our 49 Radiciotti

justifies himself by quoting what he believes to be Lowe’s evaluation of the CLA in entry nt. 417 (which, on the contrary, as we already saw before, is a work of Bischoff). Apart from the fact that the entry by Bischoff, as we already said, is contradictory and alternates two different definitions of the same script, by calling it once “Caroline” and at another time “Early Caroline,” the fact still remains that a paleographical evaluation made in 1947 and reasserted in 1965 is, in any event, old fashioned and cannot dispense us from an evaluation made with more updated standards. In fact, fundamental studies on the Caroline handwriting inspired by new methodologies were published after 1965 (Pratesi-Cherubini, L’avventura, pp. 366-370). 50 Such as, for example, the “a” that shows itself in at least three different types (one with an open curved stroke, and another with a closed curved stroke); the “g” that shows up in at least five different ways (in the shape of “5,” like an uncial; with an open curved stroke; with a closed curved stroke; compressed and reduced in size); the “x” in at least two different shapes (one with three strokes and one with two strokes); a “d” and “b” in different shapes, two of which are peculiar and opposed to each other (one with a straight vertical stroke, which is often altered in an awkward and graceless way; the other with a doubled ascender due to a cursive writing, the same that shapes the “l” into a modern “l” with an upper curved stroke; the “m” in two different ways (an Uncial and minuscule); the “e” that sometimes is high above the line, originally in italics and sometimes short, like a Half-Uncial. 51 Petrucci, Breve storia, p. 108.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 67

27-Sep-21 8:05:57 PM

68

chapter iv

attention is the nature of its complete and perfect minuscule. For the first time since the Half-Uncial, we see a very rigorous respect for the quadrilinear system, in whose central tracks the vertical ascenders originate.”52 In light of this description, the paleographical-aesthetic confusion of the witness with which we are dealing leads us to think of that jumble of non-canonical, eclectic, and disorderly handwriting found in many manuscripts from the second half of the eighth century, in which different elements overlapped and mingled with each other. Examples can be found in Italy, but also frequently in France, Switzerland, and the south of Germany. These are scripts without a name which share common characteristics inherited from Roman cursive. Because of this inheritance, they look superficially very similar, but they lack true uniformity in their appearance because there is no single standard to which they could conform.53 We are,

Plate 1 – Biblioteca Apostolica Vaticana, Pagès 1, ff. 1r, 38r (detail): it should be noted that the same letters exhibit different shapes and at times are executed in different ways (the “d,” for example, which has an unsteady stroke, or the “b,” which sometimes has a long vertical ascender extended in one single stroke and sometimes a double ascender, or the “x,” which sometimes is made with two strokes and sometimes with three). 52 Pratesi-Cherubini,L’avventura,

p. 366. Petrucci wrote authoritatively: “Fra VII e VIII-IX secolo … si adoperano diversi tipi di scrittura, alcuni vicini all’onciale e alla semionciale della tradizione ma con elementi corsivi o comunque con forme e tratteggi contrastanti con il canone classico; altri al contrario derivanti da un’intepretazione più accurata e posata della corsiva documentaria locale e perciò ricchi di legamenti e di elementi corsivi. Sono questi ultimi tipi che la tradizione designa col termine ‘precaroline’ … e che il Pratesi ha giustamente proposto di denominare più semplicemente ‘scritture altomedievali’ …” (Petrucci, Breve storia, p. 100). 53 As

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 68

27-Sep-21 8:05:59 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



69

therefore, in the realm of those confused attempts which go by the name of “pre-Caroline,” or, if one prefers, “early medieval script,” but not a type of Caroline minuscule, which, however archaic, is still recognizable on the grounds of the criteria to which we referred above. IV. In dubiis libertas Our initial claim should now be more evident: behind the respectable and even reasonable façade of the agreement of so many illustrious scholars, the old repressed shortcomings and contradictions still prevent more reflective scholars from feeling satisfied. Because of this discomfort, this disorientation, recent interpreters — although with understandable hesitation — became scholars who were less dogmatic than those who reiterated the repressed contradictions, revealing with even greater clarity the shortcomings mentioned above. They scattered, sans paraître, small but lethal question marks next to a number of unverified clichés, weakening the walls of a building that seemed impregnable. The first question mark was written by Donald Bullough in 1991. During a very passionate investigation of the role played by certain texts among authors who came together in the Carolingian court, such as the socalled Dicta Albini and the collection of formulations of the Creed present in the Pagesianus, Bullough claimed that the miscellany in the manuscript of the Marist Fathers was “a … distinctive collection of Creeds originating at the Frankish court in the (?)790s., in a book written for the (Arch)bishop Leidrat of Lyons.”54 It may seem of little importance to place a question mark on the year 790 and suggest that Leidrat’s collection could have been put together earlier than we think, but such an insinuation means, at least in an indirect way, that the manuscript itself should perhaps be dated earlier than we think. As Douglas Dales recently wrote: “Elements that were incorporated later by Alcuin were already circulating… This confirms Bullough’s wider view that much of Alcuin’s final written educational material was the culmination of many years of teaching, first at York and then at the Frankish court. It also indicates that the use of Logic was well developed before Alcuin arrived at the court of Charlemagne.”55 The second question mark was introduced alongside “Lyon” by Munk Olsen in 199756 and preserved in an update of the volumes that the scholar 54 Bullough,

Carolingian Renewal, p. 176. Alcuin. 56 Munk Olsen, Chronique des manuscrits c, p. 36. 55 Dales,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 69

27-Sep-21 8:05:59 PM

70

chapter iv

dedicated to L’étude des auteurs classiques latins57 in 2014. This apparently small trace of doubt in a scholar quite skilled in the classification and description of high-medieval manuscripts may seem of minor importance. However, it casts doubt on the narrative that the manuscript was copied while Leidrat was bishop and that it may have been destined to increase the wealth of St. Stephen’s church with texts that the bishop had brought with him from Charlemagne’s library. In other words, this doubt means denying the reconstruction made by many scholars, among whom are Delisle and Bischoff, which sees in the court library and in the culture that it embodied the main source of support for contemporary cultural institutions, a theory which has recently been criticized by more than one scholar, giving rise to quite a few heated debates.58 Without entering these debates, which would require general reflection on the role of the great medieval libraries,59 I will limit myself to the observation that, concerning the specific matter at hand, we are unsure that a process of transcription in Lyon of works originating from the court library ever existed. As we said above, the dating of the Pagesianus to a period of time after 798, proposed by Delisle and accepted by everyone, is untenable. But just as untenable is the cultural prejudice that fuels this dating. The presupposition behind this way of thinking is the unfounded belief that only Alcuin could have inspired the rediscovery of Dialectic in the Middle Ages and that, as a result, Leidrat, the disciple, must have been inspired by his master, whereas the reverse seems impossible. But what seems true is not always in fact true. The idea that our manuscript was written after Leidrat’s appointment as bishop is grounded in a series of supplementary prejudices. The first is that the manuscript was written all at once, and the second is that the two texts by Alcuin contained within it could only have been written and disseminated after a certain date. Neither prejudice holds up when weighed against the evidence. In the following pages, we will examine the manuscript in detail and show that it is not a unitary product. Nevertheless, even without considering this side of the issue, we cannot forget that the chronology of Alcuin’s texts is all but certain, and that, in any case, they were written before the appointment of Leidrat to the episcopate in 795/6: we are not obliged, therefore, to imagine a miscellany prepared by the bishop Leidrat with an eye to donating the dialectical works to Lyon, and nothing prevents us 57 Munk

Olsen, 58 Gorman, The

L’étude, p. 9. Myth, pp. 42-85; Bullough, Charlemagne’s court library revisited, in Early Medieval Europe 12 (2003), pp. 339-363, McKitterick, Charlemagne: The Formation of a European Identity, Cambridge 2008, pp. 350-368.
 59 Petrucci, Le biblioteche antiche, pp. 527-554.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 70

27-Sep-21 8:05:59 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

71

from thinking that there could have been texts in the manuscript that were copied previously. The third question mark was introduced extremely effectively by John Magee, with his characteristically subtle wit, within a hypothetical stemma of the manuscripts of Boethius’ commentary on the Aristotelian Peri hermeneias, of which the oldest extant copy is our Roman manuscript. In view of the critical edition that he is preparing, the scholar is forced to face contradictory data that makes past scholars’ claims of finding a codex opti­ mus of the text useless. Magee has to acknowledge an early contamination of the first exemplars of the work, thus preventing us from trusting a single branch of the textual tradition. This evaluation agrees with Minio Paluello’s observations but goes further, because the Italian scholar, although affirming over and over again the existence of contaminated copies, remained always faithful to the idea (inspired by Delisle’s reconstruction) that there existed a copy belonging to Alcuin from which the text of the main manuscripts was derived, although through thousands of contaminations. In order to explain this condition of almost absolute contamination, which includes the earliest extant copy of the work, the Codex Pagesianus, Magee introduces an unknown sub-archetype (on which this copy depends) into his hypothetical stemma, in addition to the alleged archetype on which the other witnesses depend. This sub-archetype is marked as unknown with a question mark. We cannot go too deeply into the core of the reconstruction proposed by Magee, which we are unable to evaluate with the necessary degree of competence (see above cc. I-III); we limit ourselves to the observation that this third question mark, although apparently of minor importance, touches a sore point and indirectly highlights a number of absurdities in the manuscript tradition. If the Pagesianus does not depend directly on the archetype or on an alleged Alcuinian copy but on an unknown manuscript which we are compelled to indicate with a question mark, then it will not be useless to ask even deeper questions about the unresolved issue raised by the oldest copy. Nor will it be useless to try to imagine its origin and its history in a new way. In order to do this, however, it is necessary to start from scratch and perform a meticulous analysis of the witness in which we are interested. This analysis will be our subject in the pages that follow. V. Codex vetustissimus The Codex Pagesianus has been described many times by scholars, but not always with the same results. Taking their contributions into account and adding something of my own, I have attempted to reconstruct the con-

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 71

27-Sep-21 8:05:59 PM

72

chapter iv

tent and codicological structure of this manuscript so that the basis of the following exposition will be clear. V.1. Structure Ours is an artificial manuscript composed of three sections, as we will see in the analysis of the fascicle arrangement. On the flyleaf recto, before Leidrat’s signature, we find the name of Arbeo of Freising († 784), employed by Leidrat (I will analyze this note of ownership later). If this signature is really an autograph, as it appears to be, it may be possible to date at least the first half of the manuscript to sometime before 784, a dating which is confirmed by the general appearance of the handwriting of the manuscript’s copyists and by the paleographical similarity of the numerous hands to the handwriting of the scribes of manuscripts coming from Freising during Arbeo’s episcopate, such as Munich, Bayerische Staatsbibliothek clm 6309 (CLA 1272). As a result, we could date the manuscript synthetically to around the last quarter of the eighth century, taking into account that it is made of different parts which were in all likelihood written between 784 and 798. The manuscript measures mm 260 × 175 and is composed of 114 ff. (as Delisle noted, the last folia were bound the wrong way, and the sequence of the pages should be as follows: 105-108, 110-113, 109 and 114). The binding of the volume is irregular (27-36 lines per folium) and is not executed uniformly (see, for example, the codicological analysis of each section of the manuscript on this point). The parchment is of low quality and full of holes and tears, as befits a typical “working manuscript” like other exemplars owned by Arbeo of Freising, such as clm 6629 and clm 6305, which are written on a similar type of parchment, full of holes around which copyists would write the text. The modern binding in brown leather with golden friezes partially preserves an older binding which was glued to the new one. There are abundant marginal notes in the text from different hands: some are attributed to Leidrat and others have been attributed to Florus of Lyon. This figure and his writing are quite well-known, and there is no need in this context to repeat what others have authoritatively established.60

60 Holtz,

La minuscule marginale, pp. 149-166.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 72

27-Sep-21 8:05:59 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

73

V.2. Writing As Bischoff showed, the manuscript is fundamentally the work of two copyists who change ink and pens but can easily be identified. We will call these main scribes “A” and “B.” The first has a very small letter size and a characteristic writing style, with letters closely attached to each other. The second has a larger letter size and is more disorderly, placing letters on the line at irregular intervals. Both of the scribes use a “pre-Caroline,” eclectic script which does not aspire to a single graphical ideal, representing the same letter with different forms. Theirs is a handwriting replete with elements in italics, ligatures, and abbreviations of which some have an Insular origin, a regular ductus, even strokes, and ascenders that are sometimes straight and sometimes badly corrected. One characteristic is a “g” similar to a ”3,” typical of Alemannic script, which at times extends above the central body of the letter and invades the space where ascenders are drawn. The copyists do not consistently separate letters and words, which are often attached to each other (Pl. 2-3). In addition to these two main hands, which alternate through many folia in the first and second sections of the manuscript (ff. 2r-106r), other copyists occasionally appear who write some lines or pages with the same ink, trying to imitate the handwriting of the other two: some examples of these hands are can be seen in ff. 56v, 87v, 103v. First, one should notice the peculiarity of the “et,” which should be in italics but is instead written with many detached strokes, imitating cursive ligature. Another peculiarity is the Insular abbreviation for “est,” represented by a small line with one dot above and another below. The line is drawn slantwise and not horizontally, as it should be. Second, one should notice the great distance between words, the excessive extension of the ascenders, and the very small central body. Finally, one should notice the very unusual appearance of some letters, such as the crested “c” (f. 103r, line 15: “accidens”). The third section of the manuscript (ff. 106v-114r) is mostly the work of an occasional copyist (copyist “C”), who uses a bright brown ink and has handwriting somewhat similar to that of the first two scribes but much more calligraphic and rounded, much more regular and orderly, with an alignment above the line which is uniform and consistent. The ascenders are often readjusted and at times have the characteristic oblique ending of the upper stroke, Insular in origin. The strokes contrast with each other; there is obvious difference between the thick and thin strokes. The letters and words are well-detached from each other, with few ligatures and abbreviations. The sign used to abbreviate “us” and “ur” is characteristic of this hand, a sort of curl in the shape of a question mark with a closed lobe.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 73

27-Sep-21 8:05:59 PM

74

chapter iv

Plate 2 – Biblioteca Apostolica Vaticana, Pagès 1, f. 2r, copyist “A”.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 74

27-Sep-21 8:06:23 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

75

Plate 3 – Biblioteca Apostolica Vaticana, Pagès 1, f. 39r, copyist “B”.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 75

27-Sep-21 8:06:39 PM

76

chapter iv

The appearance of many letters, such as “d,” “b,” “s,” and “g” — at times in the shape of “5” — bears witness to the obvious influence of Insular minuscule (Plate 4).

Plate 4 – Biblioteca Apostolica Vaticana, Pagès 1, f. 110v, copyist “C”.

Besides this occasional copyist, two other hands can be distinguished (copyist “D” and copyist “E”). They use the same type of “pre-Caroline,” eclectic script, but they detach words and letters more regularly than the first two scribes do. These two hands write the last four leaves. The first is similar to copyist “B,” while the third has larger letters and a much more disorderly arrangement on the line than the others have (Plate 5). We should not forget that the little excerpt from Boethius’ De arithmeti­ ca and Alcuin’s dedicatory verses to Charlemagne on f. 28r were added later on. It may be possible to say that the addition was made by yet another hand, using an ink of the same type as that used by the second copyist, a hand whose form of writing is similar but not identical to those of the last two copyists of the entire manuscript. It may also be possible to say that the second of the last two copyists seems to have improved his handwriting

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 76

27-Sep-21 8:06:40 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

77

Plate 5 – Biblioteca Apostolica Vaticana, Pagès 1, f. 110v, copyist “C” (details).

considerably with time, openly imitating manuscripts written in the Caroline script of the early period and substantially modifying his handwriting’s previous appearance. In fact, his writing is aligned more regularly than the others and is generally more accurate and well-composed. The scribe employs a minuscule influenced by the Caroline script, with a graceful balance between the center and vertical bodies, straight, non-modified ascenders, a consistent and graceful detachment of letters and words, and an almost complete lack of cursive ligatures (Plate 6). There are other clues that this excerpt is an addition. In the right margin, another hand, with letters that have very small bodies, has written the Tironian note symbolizing “Vacat” in order to indicate that the verses were not supposed to be copied in future transcriptions, signaling an extemporaneous addition and suggesting a “quote” of Alcuin, something “offprogram” which has the character of an encomium. Moreover, the lines on which Alcuin’s verses are written were added using a method different from the one used in the previous leaves, namely, by engraving them into the parchment forcefully on the flesh side instead of drawing the lines with a lead point as copyists “A” and “B” did. This practice is still clearly visible on several leaves of the first section (for example, a f. 27v, line 24, 25, 30, under the words “transgressione”; “contrari”; “convertimus”).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 77

27-Sep-21 8:06:48 PM

78

chapter iv

Plate 6 – Biblioteca Apostolica Vaticana, Pagès 1, f. 109v, copyist “D” (from line 14) copyist “E” (from line 14 to the end).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 78

27-Sep-21 8:07:00 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

79

As noted above, the handwriting of some of the copyists of the Pagesia­ nus recalls the disorderly handwriting, rich in cursive elements, of some scribes active in the scriptorium of Freising, a melting pot of strange writing experiments and contaminations of all kinds. For example, there is a “familiar atmosphere” between the hands of the Pagesianus and some of the hands that wrote the manuscript Munich clm 6309 (CLA 1272). This manuscript belonged to the cathedral of Freising and was written in southern Germany on rough parchment full of holes, and, like the Pagesianus, it was transcribed between the end of Arbeo’s episcopacy († 784) and the beginning of the episcopacy of his successor, Atto (Plate 8-9).

Plate 7 – Biblioteca Apostolica Vaticana, Pagès 1, f. 28r (details). In the first line, the irregular alignment of the words by copyist “B” of the pagesianus should be noted; in the second line, the irregular alignment of the hand that added Alcuin’s verses; in the third line, compare the same letter written twice, first a letter executed ungracefully and with hesitation by the main copyist, and later on, the same letter, executed carefully by the hand that wrote the added section; in the fourth line, the copyst “B” writes irregularly some words, badly aligned; in the fifth line there are the first three verses of Alcuin’s poem, written carefully, aligned on the regular line, separating letters and words, with no ligatures or abbreviations.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 79

27-Sep-21 8:07:08 PM

80

chapter iv

Plate 8 – München, Bayerische Staatsbibliothek, clm 6039, f. 14v (details, coll. 1 and 3); Biblioteca Apostolica Vaticana, Pagès 1, fol 109v, 1r, 38r (details, coll. 2, 4).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 80

27-Sep-21 8:07:15 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



81

Plate 9 – München, Bayerische Staatsbibliothek, clm 6039, f. 14v (details, coll. 1 and 3); Biblioteca Apostolica Vaticana, Pagès 1, fol 109v, 1r, 38r (details, coll. 2, 4).

Another line of reasoning supports our assumptions. If we analyze the manuscripts “copiés par la bibliothèque du Chapitre [de Lyon] sous l’impulsion de Leidrat … datables de la fin du VIIIe siècle, du dèbut ou du premier quart du IXe siècle,”61 we discover that there is not a single one that has the slightest possibility of resembling, even vaguely, the Pagesianus. The manuscripts under examination are Lyon, BM, 336 + Paris, BnF, lat. Baluze 270 (ff. 74-75),62 Lyon, BM, 599,63 Lyon, BM, 60364 + 788 (ff. 6764), Lyon, BM, 788 (ff. 23-74) + Paris, BnF, lat. 5288 (ff. 1-2),65 Lyon 466 + 61 Holtz,

Leidrat, pp. 316-320. Katalog der festländischen Handschriften des neunten Jahrhunderts (mit Ausnahme der wisigotischen), I, Laon – Paderborn, Wiesbaden 2004, n° 2546. 63 Ibid., n° 2568. 64 Ibid., n° 2572. 65 Ibid., nni 2583-2587. 62 Bischoff,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 81

27-Sep-21 8:07:25 PM

82

chapter iv

­Paris, BnF, lat. 152 (ff. 21-25),66 Lyon, BM, 608,67 Paris, BnF, lat. 11709,68 and the Laurentianus Plut. XIV, 21.69 Without entering into a detailed analysis which could not be completed here, let us recall the simple fact that all of these manuscripts are written in an almost fully developed Caroline minuscule and that some of these specimens are dated to the ninth century, at the very least within the first fifteen years of it. In these manuscripts, letters and words are well-separated, ligatures are missing, and the ascenders generally have a serif, a round, upward-moving extension (astae clavatae), a practice common in witnesses to early Caroline minuscule and not among the scribes of the Pagesianus. It is, therefore, very unclear how one could assign the manuscript of the Marist Fathers to Lyon and hold that it was copied at the request of Leidrat, since other volumes, whose transcription he did request in Lyon, exhibit quite different paleographical tendencies. V.3. Manuscript content70 I) f. 1r-v: white, with notes of ownership citing the name of Arbeo, bishop of Freising († 784) and Leidrat’s note of donation. First section of the manuscript II) ff. 2r-11r: Porphyrius, Isagoge (translatio Boethii) f. 2r: “Inc[i]p[it] Lib[er] Hisagogaru[m] Porphirii. Cum est necessa­ rium …” 66 Ibid.,

n° 2557.

67 Ibid., n° 2575. To the list of Leidrat’s manuscripts should also be added Lyon 443, which

is, however, in Half-Uncial and Visigothic script and cannot be compared to the examples in pre-Caroline minuscule or with the early Caroline, and the Rome, Vall. E 26, whose dating to the time of Leidrat is, however, “uncertain” (Holtz). 68 Lowe, Nugae Palaeographicae, I, pp. 323-324; Holtz, Leidrat, pp. 320-321 69 Zechiel-Eckes, Eine neue Arbeitshandschrift, pp. 336-370, plate 4. 70 Abbreviations used: CPL = Clavis Patrum latinorum, ed. E. Dekkers, Turnhout 1995-2010. CALMA = Compendium Auctorum Latinorum Medii Aevi (500-1500),
edds. M. Lapidge – G. C. Garfagnini – C. Leonardi – F. Santi et alii, Firenze 2000. CSLMA = Clavis des auteurs latins du Moyen Age. Territoire français 735-987,
edds. M.-H. Jullien – F. Perelman, Turnhout 1994 (Corpus Christianorum Continuatio Mediaevalis. Clavis Scriptorum Latinorum Medii Aevi. Auctores Galliae 735-987). CPPM = J. Machielsen, Clavis patristica pseudepigraphorum medii aevi, Turnhout 1990. Keefe = S. A. Keefe, A Catalogue of Works Pertaining to the Explanation of the Creed in Caro­ lingian manuscripts, Turnhout 2012 (Instrumenta patristica
et mediaevalia. Research on the Inheritance of Early and Medieval Christianity 63). Stegmüller = F. Stegmüller, Repertorium Biblicum medii aevi, Madrid 1950-80.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 82

27-Sep-21 8:07:25 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

83

f. 11r, line 2: “… traditionem. Expl[ici]t L[i]b[er] Hisagogarum Porphirii” (Aristoteles Latinus 1, 6-7, edd. L. Minio-Paluello – B. G. Dodd, Bruges – Paris 1966, pp. 3-31) Bibl.: CALMA vol. II, 4, p. 431.

III) ff. 11r-28r : Pseudo Augustinus, Categoriae decem f. 11r: “Cum omnis scientia disciplinasque artium…” f. 28r, line 5: “… erudire. De cathegoriis Aristoteles (sic ante correctionem!) explicit ab Augustino translatis” (PL 32, 1419-1440; Aristoteles latinus, I, 1-5: Categoriae, ed. L. Minio Paluello, Bruges – Paris 1961, pp. 129-175) Bibl: Aristoteles Latinus. Codices, I-II, Roma 1939-1955, n. 406; CPPM III n. 476; CPL n. 362.

IV) f. 28r: Boethius, Institutio arithmetica, excerptum, l. I, I, 1-2 f. 28r, line 6: “Est autem sapientia rerum…” f. 28r, line 13: “… actus, dispositiones, loca, tempora.” (PL 63, coll. 1079-1080; Boèce, Institution Arithmetique, ed. Y. Guillaumin, Paris 1995, p. 6,1-2)

V) f. 28r : Alcuinus, Versus ad Carolum: f. 28r, line 14: “Versus Alcuini diaconi a[d] Carolum regem Francorum. Continet iste decem naturae uerba libellus…” f. 28r, line 24: “modo mitto legendum.” (PL 101.951; Poetae latini aevi Carolini, ed. E. Dümmler, I, MGH, 1881, p. 295, Carm. 73) Bibl: CALMA, I,2, p. 146; CSLMA vol. II, pp. 16-20.

Alcuin’s verses were probably written by a hand different from that of the copyist, a fair bit of time after the transcription of the first part of the manuscript. VI) ff. 28v-30v: Alcuinus, De dialectica, excerpta, capp. XII-XIV f. 28v: “Argumentum est rei dubiae adfirmatio et constat …” f. 30r: “… ut homo minor mundus et edictum est lex annua” f. 30v: bianco (ed. PL 10, coll. 964-68) Bibl.: CALMA, I,2, p. 14; CSLMA, II, pp. 130-133.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 83

27-Sep-21 8:07:25 PM

84

chapter iv

f. 28v was initially considered to be the last fascicle of the first section of the manuscript, as shown by two clues usually located at the end of manuscripts on the leaves left blank: a note of ownership, “Leidrat,” written upside-down in the left margin, and probationes pennae of at least two different copyists on the lower margin of the leaf. Later on, the copyist of the previous pages added the excerpt from Alcuin. The verso of the last leaf was left blank. The fascicle that follows, containing Apuleius’ Peri herme­ neias, belongs to the second section and was originally the beginning of a distinct manuscript which ended at f. 106r. VII) ff. 31r-39r: Ps.-Apuleius, Peri Hermeneias (De interpretatione) f. 31r: “Incipiunt Periermeniae Apulei. Studium sapientiae quam philosophiam vocamus…” f. 39r, line 6: “… earum non potest numerus augere. Periermieniae Apulei expliciunt in quibus continentur cathegorici syllogismi” (Apulei Platonici Madaurensis Opera quae supersunt, ed. P. Thomas, III: De philosophia libri, Lipsiae 1970, pp. 76-941; Apulei Platonici Madaurensis Opera, III, ed. C. Moreschini, Lipsiae 1991) Bibl.: R. Klibansky – F. Regen, Die Handschriften der philosophischen Werke des Apuleius, pp. 152-153 (cit. at nt. 2).

Second section of the manuscript VIII) ff. 39r-106r: Boethius, In librum Aristotelis Perihermenias (De in­ terpretatione) commentarii editio prima f. 39r: “In Periermeniis Aristotelis Commentarium Boetii Incipit Liber Primus. Magna quidem libri huius apud peripaticam (sic ante correctionem!) …” f. 106r, line 23: “… edoceat secunde editionis series explicabit” (PL 64, coll. 294-391; Boethius, In Librum Aristotelis Peri Ermeneias pars prior, ed. C. Meiser, Lipsiae 1877). Bibl.: CPL n. 883; CALMA, II, 4 pp. 428-429.

Boethius’ text ends on f. 106r, the last leaf of the section, as shown by the note “scribe usque hunc” in the lower margin and a note of ownership, “Leidrat,” on the right margin, erased but still legible under ultraviolet light, a type of note which is usually appended to the recto or verso of the last leaf of a manuscript. The texts that follow f. 106r, written by other hands, were added later on leaves ruled in a different way.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 84

27-Sep-21 8:07:25 PM

85

Arbeo, Leidrat, Alcuin, and the codex pagesianus



IX) Third section of the manuscript f. 106v-107r: Pseudo Alcuinus, Dicta Albini diaconi de imagine Dei f. 106v: “Faciam[us] hominem ad imaginem et similitudinem n[ost]ram.”  

f. 107r, line 15: “in primo Adam condidit, mirabiliusque in secundo reformavit.” (PL 17, coll. 1105-1108; cfr. PL 40, coll. 805-806; J. Marenbon, From the Circle of Alcuin to the School of Auxerre: Logic, Theology and Philosophy in the Early Middle Ages, Cambridge – London – New York – New Rochelle Melbourne – Sidney 2006 [Cambridge Studies in Medieval Life and Thoughts, Third Series, 15], pp. 158-161) Bibl.: CPPLMA, II B n. 3008a; CSLMA, II, p. 520.

This text and the others up to ff. 114r-v were bundled together at a later time, collected into a poorly bound and poorly numbered fascicle which swapped the positions of some of the leaves. X) ff. 107r-107v: Credo Nicenum f. 107r, line 25: “Exemplum Fidei Niceni Consilii CCCXVIII E[pisco­ ­p]orum, credim[us] in unu[m] d[eu]m patrem om[ni]p[otentem]  

107v, line 5: “… hos anathematizat catholica et apostolica ecclesia” (The text is quoted by Rufinus Aquileiensis, Historia Ecclesiastica, I, 6, PL 21, coll. 472-73)

XI) f. 107v: Pseudo Ambrosius, Fides Sancti Ambrosii Episcopi f. 107v, line 5: “Item Fides S[an]c[t]i Ambrosii Ep[iscop]i. Nos Patrem et Filium et Sp[iritu]m s[an]c[tu]m confitemur, ita ut in trinitate…” f. 107r, line 12: “… ideo alienus est ac profanus et adversus veritatem rebellis” (Manuscripts canonum ecclesiasticorum et constitutionum sanctae Sedis apostolicae, c. 37, PL 56, coll. 582-83) Bibl.: CPL, n. 789; CPPM, II, A, n. 36a; Keefe, pp. 133-134, n. 212.

XII) f. 107v: Gregorius Turonensis, Fides Gregorii f. 107v, line 13: “Incipit Fides S[an]c[t]i Gregorii Papae Urbis Romae: Credo d[omi]n[u]m patrem om[ni]p[otentem], credo I[esu]m [Christu]m filiu[m] ei[us] unicu[m]…” f. 107v, line 32: “… et omnia quae a trecentus XVIII episcopis niceni concilii congregatis in statuta sunt credo fideliter…” (Gregorius Turonensis, Historia ecclesiastica Francorum, Prologus, PL 71, coll. 159-63).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 85

27-Sep-21 8:07:25 PM

86

chapter iv

XIII) f. 108r: Gregorius Thaumaturgus Expositio fidei (translatio Rufini) f. 108r: “Item Fides Beati Gregorii Martyris et Ep[iscop]i Neo Necessa­ riae. Unus d[eu]s pat[er] uerbi uiuentis sapientiae subsistentis …” f. 108r, line 14: “… sed inconuertibilis et immutabilis eadem trinitas semp[er] manet.” (PG 10, coll. 983-87)

XIV) ff. 108r-v and 110r: Pelagius monachus, Libellus fidei ad Innocen­ tium papam f. 108r, line 15: “Expositio Fidei catholicae S[an]c[t]i Hieronimi. Cre­ dimus in d[eu]m patrem omnipotentem, cunctor[um] visibiliu[m] et invisibiliu[m] conditorem …” f. 108v, line 32: “… secundum deitatem suam.” The text continued onto the next leaf, starting with the words “impossibilis est,” but the leaf was badly bound and currently has the number 110r instead of 109r. The text continues on the same leaf and ends on f. 110r, line 31 with the words “…non me hereticum conprobauit.” (ff. 109r-109v were inserted between f. 108 and f. 110 by mistake, whereas they should have been placed after them). (PL 45, coll. 1716-18) Bibl.: CPL n. 633 and 731; CPPM, I, nn. 1021, 5007; CPPM, II, A, nn. 205, 865, 1392; Keefe, pp. 75-76, n. 39.

XV) f. 110v: Iuvencus, Evangeliorum libri IV, excerptum, L. I, vv. 590603. f. 110v, line 1: “Siderio genitor residens in uertice caeli…” f. 110v, line 14: “… cedere nec durum erratis intendere pectus.” (edd. PL 19, coll. 132-133; G. V. A. Iuvenci Evangeliorum libri IV, ed. I. Hueher, Pragae – Vindobonae – Lipsiae 1891 [CSEL 24], p. 32).

XVI) ff. 110v-112v: Cassiodorus, Expositio Psalmorum, Praefatio, cc. IIII. f. 110v, line 14: “De Prophetia, Prophetia est aspiracio divina…” The text, written by copyist D, runs until the words “… vocem de coelo” at line 29, f. 110v. It then starts again, this time written by copyist E, from the words “nec per alios” on the first line of f. 111r until f. 112v, line 20: “… modolationes musicas exprimentes.” (PL 70.12-15; Cassiodorus, Expositio Psalmorum, ed. M. Adriaen, I, Turnhout 1958 (CCSL, 97), pp. 7-12; Cassiodorus, Expositio Psalmorum, ed. P. Stoppacci, Firenze 2012, pp. 379-410)

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 86

27-Sep-21 8:07:25 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



87

Bibl.: CALMA vol. II, 5, pp. 576-577; CCSL vol. XCVII-XCVIII; CPL n. 900; Stegmüller, n. 1894.

XVII) ff. 112v-113r: Pseudo Damasus, Epistola ad Hieronymum f. 112v, line 20: “In [Ch]r[ist]i Nomine. Incipit epistolae papae Damasi ad Hieronimum presbiterum: Damasus ep[iscopus] f[rat]ri et conpresbitero Hieronimo in [Christ]o salutem. Du[m] multa corpora librorum in meo arbitrio…” f. 113 r, line 2: “… P[er] Bonifatium presbiterum Hierusolimam.” (PL 13, 440-41 = Hyeronimus, Epistolae suppositae, 46, PL 30, 294-95).

XVIII) f. 113r: Pseudo Hieronymus, Ep. XLVII ad Damasum f. 113r, line 4: “Item S[an]c[t]i Hieronimi ad Damasum. Beatissimo papa Damaso sedis apostolicae urbis Romae Hieronimus supplex. Legi litteras apostolatus…” f. 113, line 22: “… quod est alleluia, quod grece dicitur prologus latine dicitur praefacio.” (Hyeronimus, Epistolae suppositae 47, PL 30, 294-95). Bibl.: CPL n. 633

XIX) f. 113v: Pseudo Hyeronimus, De Psalterio (Prologus) f. 113r, line 22: “Item alia epistula unde supra. Nunc au[tem] exposuimus originem omnium psalmorum et nunc…” f. 113r, line 28: “… et omnis sp[iritu]s laudet d[omi]n[u]m» (PL 30, 296, republished without the author’s name in PL, Supplementum, 274-75); D. De Bruyne, Préfaces de la Bible latine, Namur 1920, pp. 44-45, nn. 2-3)

XX) f. 113v and 109r: Pseudo Hyeronimus, Versus f. 113 v, line 29: “Item unde supra, Damasi et Hieronimi. Psallere qui docuit dulci modulamine s[an]c[t]is…” f. 113v, line 36: “… ut d[omi]no” concludes with the words “… et quos gratiae vocis” on f. 109 r, lin 1, falsely inserted at this point. (PL 13, 375, frammento 1).

XXI) f. 109r: Pseudo Damasus, Carmen I f. 109r, line 1: “Nunc Damasi monitis aures p[rae]bete benignas …” f. 109r, line 12: “Hec Damasus scit s[an]c[t]e tuos ipse triumphos.” (PL 13, 375).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 87

27-Sep-21 8:07:25 PM

88

chapter iv

XXII) f. 109r: Isidorus Hispalensis, In libros veteris ac novi Testamenti prooemia, Proemium de psalterio, cc. 33-35 f. 109r, line 12: “Item de libro s[an]c[t]i Isidori. Liber psalmorum quamquam uno concludatur volumine…” f. 109r, line 29: “… ordinauit ut voluit” (PL 83, 163-64) Bibl.: CPL n. 1192; Stegmüller, nn. 5176-5231)

XXIII) f. 109r-v: Pseudo Augustinus, Epistola ad Paulam et Eustochium de utilitate psalmorum, excerptum f. 109r, line 29: “S[an]c[tu]s Augustinus dixit canticum psalmorum an mas decorat invitat angelos…” f. 109v, line 14: “… in celo mirificabit in s[e]c[u]la s[e]c[u]lorum amen” (Quoted in in Remigius of Auxerre, Enarrationes in Psalmos, PL 131.142. The text is a Latin readaptation of Basilius Caesarensis, Homilia in Psalmus I, PG 29.210-14). Bibl.: CPPLMA, II, A, nn. 1907, 2363.

XXIV) f. 109v and 114r-v: Credo Atanasianum f. 109v, line 14: “Fides s[an]c[t]i Athanasii ep[iscop]i alexandrini. Quicumq[ue] vult salvus esse ante om[ni]a opus est…” f. 109 v, line 26: “… inmensus et sp[iritu] s[an]c[tu]s, aeternus pat[er]” the text continues at line 1 of f. 114 r, which ought originally to have followed the preceding one, with the words: “… eternus filius, eternus s[an]c[tu]s sp[iritu]s …” and ends at f. 114v, line 31, with the words: “… saluus esse non poterit”  

(PL 88, col. 585. The same text in Greek is in PG 28, col. 1581).

V.4. Fascicle arrangement First Section: Fascicles: 4 + 4; 4 + 4; 3 + 3; 4 + 4. The present fascicle has been rearranged: originally, there were six leaves to which two leaves were added so as to form a seemingly regular fascicle made of 4 + 4 leaves like the previous ones. This section of the manuscript first belonged to bishop Arbeo and later to Leidrat, as shown by their consecutive notes of ownership on f. 1r and 28v. The fascicles are numbered from I to IV with Roman numerals in the middle of the lower margin on the last leaf of each fascicle. The first three fascicles are numbered by a hand that is unlike the one that numbered

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 88

27-Sep-21 8:07:25 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

89

the last fascicle and the other fascicles of the second section. In order to understand this, it may suffice to compare the ascenders of the numbers in the first three fascicles with those of the other fascicles. To that end, we should also compare the two wavy lines that come together, surrounding the numbers, with the very short and scattered lines delineating the numbers in the second section. The first hand draws two very long and winding lines around the numbers, culminating in a point and forming some sort of triangle. The second hand, which only marks the number IV, writes numerals with a slimmer body and draws a little angle above them, forming two very short lines without any winding ending (it is probably the scribe of the second section who sketches the same angle in the second part of the manuscript: see, for example, the number in the lower margin of f. 48r). The fascicle begins on the flesh side, and the leaves are ruled using a double lateral line inlaid in both the external and internal margins, at times on four bifolia, by pressing on the flesh side with a pointed tool. Some internal leaves in the fascicle have two small lateral holes that help with drawing lines above and below (f. 3r-v), or a series of small holes (at least 16), pressed in-between the double line on the right of the external margin. Generally, the small holes show up consecutively on different leaves (4-5; 6-7-8; 19-20, 21-22). In the third fascicle, the pointed tool is pressed on the leaf in vertically (f. 22r-v, sixth leaf of the second fascicle). In the fourth fascicle, the small holes were inlaid by holding the pointed tool horizontally (f. 25r-v, third leaf of the fourth fascicle). There are traces of at least two separate tools for making the small holes (on ff. 12r and 15r). The columns of the small holes are intact and run along the margin, with one exception: the small holes following Alcuin’s verses on f. 28r. They were drawn with the express purpose of sketching out the lines within which the verses were to be written and are sufficient only to draw those lines. Moreover, the small holes are arranged so that they are attached to the last words of each verse; they are not on the outermost margin, as is the case in the remainder of the manuscript. The double lateral line found in the previous leaves is missing. The work was executed so badly that three or four small holes, not properly aligned, can be seen to have invaded the space above the verses and to have done so without apparent reason, for they do not correspond to the lines of the fragment of Boethius’ De arith­ metica. This is the copyist’s mistake. He measured incorrectly at the beginning and then made the small holes out-of-line and in the wrong place. Later on, realizing his mistake, the copyist corrected himself and made the small holes lower down, arranging them vertically so that he could draw the necessary lines. In the outer and lower margins of the fascicles it is still possible to dis-

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 89

27-Sep-21 8:07:25 PM

90

chapter iv

cern the marks of someone supervising the copying. He writes expressions such as “usque huc,” or simply “usque” with a dry point and without ink in order to show where the text copied by the scribe was supposed to end, a practice attested in other manuscripts. The same type of marks are also found farther on in the second section of the manuscript (for example, on f. 83r). Initially, f. 28v was blank and was considered the last fascicle of the first section of the manuscript, as indicated by two clues which are normally found at the end of manuscripts on the leaves which have been left blank: an autographic note of ownership, “Leidrad,” written upside-down in the left margin, and probationes pennae of at least two different copyists (one wrote “pal”; the other first wrote: “dei f[ilius] [et] [vi]rgo” and, afterwards, “virgo”) on the lower margin of the leaf. Later on, the blank space was filled by transcribing the beginning of the Alcuinian extract on the blank side of the page and continuing onto two leaves that were then glued to the fascicle. In this way, it was transformed from a sexternion into a quaternion and numbered by a hand different from the one that numbered the previous fascicles. Nevertheless, one side of the page at f. 30v was left blank for no apparent reason, indicating the negligence and inattention of the person who added the leaves to the fascicle. Second section: ff. 31r-106v Fascicles: 5 + 5; 4 + 4; 4 + 4; 4 + 4; 4 + 4; 4 + 4; 4 + 4; 4 + 4; 4 + 4; + 2 ff. Boethius’s text ends on f. 106r, the last written leaf, as indicated by the note “scribe usque hunc” in the lower margin and the note of ownership, “Leidrat,” that was erased but is still legible under ultraviolet, a note which is usually supposed to be on the last leaf of a manuscript or on the first recto or verso. The second section has a fascicle numbering that is more irregular than the first and uses a slightly different ruling. In addition to double lines drawn in the margins, we find another method of ruling: creating small holes between the two lines, but less systematically than in the first part. In fact, the small holes appear inconsistently. Sometimes, there is a column of small holes on only one or two pages (ff. 61r-v; 67r-v and 68r-v; 80r-v; 84r-v and 85r-v; 93r-v; 104r-v). Sometimes there are two columns (f. 45r-v; 5r-v). Sometimes there isn’t even an entire column (e.g. ff. 79r‑v; 88r‑v). And sometimes there are three holes in the middle (e.g. ff. 60r; 62r). In any event, the holes are positioned with some irregularity, made with tools with different tips and arranged on the leaf sometimes horizontally and sometimes vertically (e.g. ff. 85r-v; 45r-v). Each fascicle is numbered in the lower margin with roman numerals,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 90

27-Sep-21 8:07:26 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

91

starting with the number V in order to preserve the numbering of the previous section. The hand that wrote these numbers is that of copyist “A” (we see on f. 40v, in the left margin, the number “V” followed by a dot on the right and one on the left, and then the number itself, drawn in an identical way with the same lateral dots). Nevertheless, the fascicles were originally marked differently, with dry-point ruling or by writing the first word of the following fascicle in very small characters on the last leaf of the fascicle using the tip of the pen. Traces can be found on ff. 48v (“ea[dem]: the first word of the fascicle starting with 49r), 88v (“sunt”: the first word of the fascicle starting on 89r), and ([eni]m ver[o]: the first two words of the fascicle starting on 105r). Additionally, in the first section, we find the first word of the fascicle starting on 27r, “neque,” written with a dry point on f. 26v. All of these clues lead us to think that the two sections were originally separate and that only afterwards did someone decide to put them together into a single corpus. It is no coincidence that the fasciculation and the binding in the two sections are so different, as if the work of the copyists had been guided by different criteria. It is also no coincidence that in both of them we find the repeated note of ownership, “Leidrat,” on what was supposed to be the final leaf. Still, the presence of copyist “B” and of the same inks in the first and second sections assures us that the transcriptions of the two sections were performed under the same circumstances, with one following not long after the other. Third section: ff. 107r-114v Fascicles: 2 ff.; 3 + 3 ff. (at present, the final fascicle is a reconstruction of separate leaves, assembled after the transcription of the manuscript, and in a way that is different from the original). Different ruling systems alternate in the leaves (a single lateral line on ff. 110 and 112; double lines on ff. 106, 107, 108; only small holes elsewhere). There are also different ways of designating the lines, sometimes marking the parchment clearly, other times not. Initially, ff. 107-109 were the only added leaves, leaves which nevertheless correspond to the original f. 113. The correct sequence of leaves is the following: ff. 107, 108, 113 (orig. 109), 110, 111, 112, 109 (orig. 113). We can reconstruct the correct order by examining the texts contained in them, following each other coherently only in this order. We should add that some probationes pennae71 are written on what is now f. 109v (originally 113v), as is customary in the final blank space of the manuscript, a 71 The text of most of these annotations was transcribed in Burn, Facsimiles; Bischoff, Mittelalterliche, 3, p. 122.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 91

27-Sep-21 8:07:26 PM

92

chapter iv

space that, to tell the truth, had already begun on f. 109r (originally 113r), where there were some unwritten lines. A passage was added later that begins in the section left blank on f. 109r, continuing onto f. 109v (originally 113v) and ending at f. 114r, written around the annotations which already existed. It is clear that the addition was made after the probationes pennae were appended. One of these “tests” of the copyist is significant: in fact, one hand that intervened later in order to correct the text on f. 109v, line 18, jotted down the signature: “Arvitus subscripsi.” Contrary to what one may think, this subscriptio, accompanied by the written ticks that we usually find in an authentic document, does not indicate the scribe’s name in a strange way, but is rather a probatio penna borrowing a word from another context. It could be a signature of the Acts of the Council of Toledo of 589, where the bishop of Porto was present, a man with the very rare name of Arvitus.72 He signed with these exact words: “Arvitus scripsi.” This means that the copyist had the Acts of this council in his hands, which, as noted by Dales,73 was of great importance for the Antiadoptionist controversy and was used by those like Leidrat who argued against the Spanish heretics. As a result, we can state that the last two leaves of the manuscript (but only those and no others) — what are now ff. 109 and 114 and were originally ff. 113 and 114 — were added to the others slightly later, perhaps after the missions in Spain between 797 and 799. Notae possessoris On leaf 1 it is possible to read a half-erased note of ownership written in a pre-Caroline calligraphic script, graceful and accurate, with a flowing and consistent stroke. This handwriting exhibits a harmony of proportion between the central body and the ascenders, a characteristic of early Caroline script, but there are still some “pre-Caroline” elements, such as some cursive ligatures and letters that were originally cursive (as, for example, the “e” above the line). The handwriting also shows evident Insular influence (as can be seen, for example, in the “e,” “p,” and “b”). It shows signs of contact with Leidrat’s handwriting and characteristics which are somewhat similar.74 However, it has a less cursive format and, as a consequence, does not excessively elongate the stems of letters ending above or below the line, stems which often feature serifs and extensions of the strokes 72 Concilia

Toletana III, p. ccxxxv. Alcuin, p. 72. 74 Such as the stroke of “d”, “p,” and the conjunction “et” (within the word), with the stroke indicating that the “t” is extremely prolonged. 73 Dales,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 92

27-Sep-21 8:07:26 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

93

(apart from “s” at the end of a word, and the final stroke which forms the conjunction “et”). The note consists of a quotation from Isaiah 59:5 and bishop Arbeo’s name (764-84): “Et quod confotum est erumpet [in] regulum” (Is. 59, 5). ‫٭‬ S iG (= Signum75) Arbeei» (Pl. 10-11).

Plate 10 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1r (detail): The name Arbeo is written “Ar beei”, after the abbreviation “si” meaning “signum”. 75 This is a very common abbreviation in the signatures of documents: cfr. Cappelli, Dizionario, p. 351, 1 col.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 93

27-Sep-21 8:07:30 PM

94

chapter iv

Plate 11 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1r: Arbeo’s note of ownership.

The formula which we find in the signature is the sort which was typically used from the 6th c. onwards (Papiri Marini 90; 114), in these cases as an alternative to “Ego … subscripsi.” The bishop Arbeo uses it at least twice, signing two documents with the formulae “Signum Arbionis episcopi” and “Signum manus Arbioni episcopi.”76 The hand that jots down the note of ownership occurs in the Pagesianus in the left margin of ff. 17r and 19r, in different annotations that eventually subdivide the paragraphs of a meaningful passage from the pseudo-Augustinian Categoriae77 with headings. The same writing style occurs

Plate 12 – The first three annotations were written in the margins of Lyon, BM, 402 on the leaves 22r and 22v. The fourth is the signature on f. 1r of the pagesianus 76 «Signum manus Arbioni episcopi qui interfuerit» (doc 25: Petenbrunn 765-770, codice A 38v, n. 17); «Signum Arbionis episcopi» (doc 28: Kronacher 26 aprile 768, Cod. A f. 38, n. 16), edited in Bitterauf, Die Traditionen des Hochstifts, pp. 55-56. 77 The passage corresponds to paragraphs 71-77, pp. 149-150 of the Minio Paluello edition, quoted in n. 5.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 94

27-Sep-21 8:07:35 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



95

in the margins and blank spaces of other manuscripts belonging to Leidrat, mixed with other notes by the same Leidrat, such as those in clm 6305 (CLA 1269) of the Bayerische Staatsbibliothek in Munich,78 originating in the scriptorium of the cathedral of Freising, or those in Lyon 402 (CLA 770), of uncertain origin (Pl. 12-13).79

Plate 13 – Comparison of some letters from the annotations in Lyon, BM, 402 (first column, f. 2r) and Munich, Bayerische Staatsbibliothek, CLM 6305 (first column, ff. 10r, 30r) with Arbeos’s signature of the Pagesianus manuscript (second column, f. 1r) 78 Glauche, 79 Bischoff,

Katalog der lateinischen Handschriften (cit. at nt. 36), p. 184. Katalog, n° 2547b.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 95

27-Sep-21 8:07:36 PM

96

chapter iv

Plate 14 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1r (detail).

On the left side of the owner’s note in Pagès 1, a rather rough hand, almost erased, and not long after the time of the first note, has ruled some upper-case and lower-case letters with a dry point, arranging them quite far apart in a vertical orientation: the letters are “Ar B e o”80 (Plate 14). On f. 1v we find the signature containing Leidrat’s donation to the church of St. Stephen of Lyon (to which we will return later on). Under the hand is written upside-down and by a slightly later hand “Atque in [h]unc modo contendeb[ant]” (Reg., III, 22). A little lower, a ninth-century hand wrote: “Arbeei.” The two “e”s of the name are corrected in the line space by writing one single “e” so as to have the name in its most common orthographic form, namely “Arbei” and not “Arbeei.” The note of ownership was later erased, having been corrected with ink (Plate 15). If the different notes of ownership really mean that the first part of the manuscript belonged to bishop Arbeo, it follows that the high-medieval rediscovery of the pseudo-Augustinian Categoriae is to be attributed to the bishop of Freising and not to Alcuin, as is often claimed. It is no accident — as we saw before — that the verses by Alcuin dedicating this little work to Charlemagne were added after the manuscript was transcribed. It is also not a coincidence — precisely because they were added later on — that 80 Perhaps it is still possible to read “hoc F” (with a sign of horizontal abbreviation), which could mean “hoc fecit”

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 96

27-Sep-21 8:07:36 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

97

Plate 15 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1r (detail).

they are at the end of the text, not at the beginning or in the introductory format according to which they should have been arranged and actually are arranged in the majority of the manuscripts which contain them. In this case, therefore, it may be presumed that Leidrat had Alcuin’s verses added while at the Carolingian court. By accepting this version of events, we must imagine that Alcuin “offered” the booklet of the pseudo-Augustinian Categoriae as a gift to Charlemagne, and he offered it only in the sense that he physically gave to the king a copy of a text that was obtained elsewhere (perhaps through Leidrat himself). The possibility that the “rediscovery” of logical works in the Middle Ages should be linked to Arbeo’s name is neither strange nor impossible.81 It was a fully conscious rediscovery, if we know how to interpret the subtle allusion in the quotation from Isaiah that accompanies the name of the bishop of Freising in the note of ownership. The image of a snake unex81 Concerning

the manuscript produced in Freising at the time of Arbeo and generally concerning the library attached to the Dome of Freising, see for a first introduction Bischoff, Die südostdeutschen Schreibschulen, I, pp. 60-68 (cit. at nt. 29); Glauche, Bistum Freising, pp. 617-618. See also Maag, Alemannische Minuskel. About Arbeo, see Arbeo von Freising, col. 888; Vita Corbiniani.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 97

27-Sep-21 8:07:43 PM

98

chapter iv

pectedly crawling out of an egg that looks like any other egg is obviously related to the image of the snake at the end of the Dialectica, an allegorical representation which was often employed in the Middle Ages and that is already present in Martianus Capella, who describes Dialectic holding its “viperish” snakes82 in its left hand. These new ideas came from old books: the manuscripts which men like Arbeo, born in Merano, could obtain in northern Italy83 from its very old and indomitable libraries of Late Antiquity, ones that survived destruction with the cunning of the snake and the prudence of the turtledove, libraries like the one in Verona which housed manuscripts from the time of Boethius, or like the one in Bobbio, where there were Irish monks such as Virgil of Salzburg and Peregrinus who, as we said, worked on behalf of the bishop and transcribed several manuscripts for the cathedral of Fresing. VI. Leidrat’s annotation As noted before, many book dedications, corrections, and annotations are confusedly and inconsistently attributed to Leidrat. It is necessary, therefore, to shed light on his writing, assigning to the bishop of Lyon what really belongs to him and eliminating what does not.84 Concerning the dedications of the volumes in which his name occurs, a simple comparison shows that at least four out of the six signatures were written by Leidrat. Only two, that in the Pagesianus and that in Lyon 608, were written by the same person and show the same handwriting. We find it again in marginal notes in the manuscripts of this tireless reader (Plate 16).

Plate 16a – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v: The signature by Leidrat. 82 “Laeva sub pallio occulebat insidias viperinas, cunctis dextera praebebatur” M. Capella, De Nuptis Philologiae et Mercurii, IV, 423, p. 99. 83 About the circulation of books and ideas in North Italy, Salzburg and Freising, see for a first introduction to the issue Le Alpi porta d’Europa 84 On this topic, see the recent intervention by Holtz, Leidrat, évêque de Lyon.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 98

27-Sep-21 8:07:47 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



99

Plate 16a – Leidrat’s subscriptions in the manuscripts: Biblioteca Apostolica Vaticana, Pagès 1, f. 1v; Lyon, BM, 608, f. 1r; Paris BnF, lat. 152, f. 25r; Lyon, BM, N, 599, f. 1r; Paris BnF, lat. 11709, f. 1r.

The signature in Paris, BnF, lat. 11709 is a bad imitation of Merovingian minuscule which has nothing to do with the cursive script of the Page­ sianus; that it was a false signature was already maintained by Holtz.85 We should note the difference of the “L” and, above all, of the “d,” which in the Parisian witness is written with a “crested” lobe. This practice, besides being contrary to the style of Merovingian script, has nothing to do with the lobe in two strokes of the Pagesianus or with the one belonging to other signatures. Additionally, the subscriptio in Paris, BnF, lat. 152 is decidedly different from the others. The “L,” in this case, has a low, bent stroke, unlike other examples in which it contains a right angle. The “r” definitely descends be­ low the line, unlike the other examples in which it is always on the line. The “d” ends with a small line completing it, a line that is oddly in the shape of a club, whereas it is straight in the other examples. The space occupied by the name of the bishop is tightly squeezed, compressed, and reduced, whereas the other signatures usually include ample space between the letters. 85 Holtz,

Leidrat, p. 320.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 99

27-Sep-21 8:07:55 PM

100

chapter iv

Concerning the subscriptio of Lyon 599, as we recalled above, it was not considered an autograph by Bischoff, and we can only agree with his judgement. We have to reconsider, therefore, the actual number of Leidrat’s autograph dedications. But this comes as no surprise. Delisle had already observed that there was a ninth-century dedication in Lyon 610 which was spurious, and he did not attribute any particular meaning to this fact, interpreting it as an attempt to replace the original lost signature. In turn, as we just said, Bischoff pointed out that the dedication of Lyon 599 is not the work of the bishop of Lyon but was added to the manuscript by other people. It cannot come as a surprise, then, that other manuscripts should exhibit the same phenomenon and that the signatures are false. We should limit ourselves, therefore, to considering only two manuscripts if we wish to describe Leidrat’s handwriting with a full awareness of the facts. The bishop wrote with a pen that has a very fine point, typical of documentary cursive script in Caroline minuscule strongly influenced by cursive, with the vertical stems significantly extended and the main body of the letters extremely small, alternating between letters that are definitively cursive (such as the extended “e” or the crested “C”) with letters which have a more even appearance. The same handwriting is found in annotations in the margins and in the line spaces of the Pagesianus, clm 6305, and Lyon 402. Other scholars have pointed this fact out, but in such a way that they did not distinguish carefully between Leidrat’s hand and those of other correctors of the same manuscripts (Plate 17).

Plate 17 – Biblioteca Apostolica Vaticana, Pagès 1, f. 113r (detail); München, Bayerische Staats­bibliothek, CLM 6305, foll 3r, 4v, 5r (details); Lyon, BM, 402, fol 18r (detail).

As Louis Holtz authoritatively wrote, “la longueur des hastes, surtout des hastes supérieurs est bien visible dans la formule de dédicace, mais on

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 100

27-Sep-21 8:08:01 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



101

trouve aussi des ligatures spécifiques, celle de «us» … et de «en». La graphie du «e» est particulièrment caracteristique: sur une base réssemblant au debut d’un «m», veint se greffer une boucle supérieure, frêle et comme superposèe.”86 If we accept this description as accurate, we can no longer claim that the majority of the corrections in the Pagesianus are by Leidrat’s hand,87 nor can we state that, in other manuscripts in which we find traces of his intervention, he was the only one to have corrected the texts. VII. Recapitulatio On the basis of what we have observed, we have to conclude that Pagès 1 consists of different parts brought together after their composition. The first part was surely transcribed before Arbeo’s death, because the bishop added the note of ownership on f. 1r. It must be dated before 784. This section belonged to Leidrat after Arbeo’s death, as shown by his note of ownership on f. 28v. In this section are the pseudo-Augustinian Categoriae, which, according to general opinion, were “discovered” by Alcuin and then at a later time passed on to Leidrat. But the opposite is true. Concerning the dedicatory verses to Charlemagne, written around 795-96, as we said before, they were added much later, after Arbeo’s death, as was the abstract of the Alcuinian De dialectica. The second section of the volume was mostly copied by the scribes who completed the first section and is most likely contemporary with it or shortly posterior to it. Besides, considering the types of works contained within it, it is likely that it stemmed from interests common to both Arbeo and Leidrat. The latter, in any event, left a note of ownership on f. 106r and, as a result, he was the owner of this section of the manuscript when it was still separate from the first section. As we have outlined the Libri Carolini, composed between 790 and 795, quoted some passages from Boethius’ texts extracted from the Pagès 1 (cfr. infra, p. 116). The union of the two sections must have happened later on. It was carried out by the copyist of the first section, who numbered the fascicles in sequence, linking them to the numbering begun by a different hand in the first section. It is probable that Leidrat himself asked for the recombination of the two volumes which he owned, perhaps during his residence at the Carolingian court. The third part of the volume, comprising texts meaningful to the intellectual Carolingian élite that produced the Libri Carolini (which the texts themselves quote and rework) was presumably transcribed during the pe86 Holtz, 87 Ward,

Leidrat, p. 326 (cit. at nt. 5). Pagesianus, p. 310.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 101

27-Sep-21 8:08:01 PM

102

chapter iv

riod in which the Libri were composed, between 790 and 793. In any event, as we saw before, the last two leaves were added shortly before or shortly after the period from 797 and 799. The timespan within which we must date the manuscript is 784-798, taking into account, however, that the different parts of which it is composed, although transcribed not long after each other, were not completed during the same years. VIII. Legi There remains in the Codex Pagesianus a fascinating enigma. From f. 28r until f. 48r, we find many interlinear and marginal corrections of Boethius’s Commentary on the Peri hermeneias, written by a hand contemporary with that of the copyist.88 The corrector’s hand employs a type of ink that is dimmer than the one used by the scribe and writes with a graceful minuscule, very calligraphic, essentially Caroline, but with a marked Insular influence. This influence is visible both in the characteristic forms of letters such as “l” and “d” and in typical abbreviations, such as the one for “que,” represented by a “q” followed by three dots in the shape of a tri. angle (f. 4r, line 13 = q . .) (Plate 18).

Plate 18 – Biblioteca Apostolica Vaticana, Pagès 1, ff. 38r, 41r. 88 The

hand is clearly anterior to the annotations attributed to Floro of Lyon (800-860), who erases his corrections in order to insert new ones as alternatives (e.g. f. 74r, line 24 and right margin).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 102

27-Sep-21 8:08:04 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

103

It is an early Caroline script (this is the proper time to say it!), composed between the last quarter of the eighth and the beginning of the ninth century, showing affinities with the eclectic handwriting developed in the French centers in which Insular influence was strong. We can associate our corrections with one of the hands at work in the manuscript Florence Laur. Plut. XLV, 15, written in Tours in the second half of the 8th c. (Plate 19).

Plate 19 – Firenze, Biblioteca Medicea Laurenziana, Plut. XLV, 15, ff. 14r, 17r, 19v, 45r, (first column, details); Biblioteca Apostolica Vaticana, Pagès 1, ff. 38r, 47r (second column, details).

Whoever corrected the Pagesianus was assisted in his task by a collaborator using a Caroline script full of Insular elements. The assistant’s writing is more disorderly and uncertain, with a stroke that is not always

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 103

27-Sep-21 8:08:09 PM

104

chapter iv

smooth, as shown by his alterations to stems and the breaking of some curvatures (Plate 20).

Plate 20 – Biblioteca Apostolica Vaticana, Pagès 1, f. 38r (detail).

The first hand, assisted by the second, allows itself the luxury — a rare privilege at that time — of correcting the text of Boethius’s Commentary with the help of another exemplar, enabling the insertion of some lines skipped in the first transcription of the text (f. 28r, right marg.). Nevertheless, the systematic correction is interrupted about one-third of the way through the commentary, the last lines of f. 48r. The hand under examination introduces only occasional corrections from then on, emending the

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 104

27-Sep-21 8:08:14 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

105

distractions and orthographic mistakes that any careful reader would have discovered, even without the help of an antigraph for checking the text. This change suggests that the corrector had a partial text of the Commen­ tary available to him, limited to its first chapters, and that he read the section of the Pagesianus which he did not know with profound interest and care, intervening in the text only when he felt he could do it with the help of his critical skills. Whoever carried out the correction, however, had a deep interest in what he was reading, an interest attested by many marginal markings of passages that are useful for quotation or extraction. The marks in the margins were made by appending an “S,” meaning “scribe,” that is, “extract the text” in the margins of the marked text, sometimes indicating the point at which the quotation should end with a strong period (cfr. above, p. 30). Who is this corrector writing in such a peculiar way, who has in his hands a fragmentary manuscript, so rare, and, above all, who has the authority to amend Leidrat’s specially-made and exclusive manuscript, who reported texts that no one else knew until that moment? An identification of this unquestionably extraordinary reader may be made by examining a signature containing the initials of an abbreviated name written with the same pale brown ink used for the corrections. It is found in the lower margin of the first leaf on which he started to revise the text, f. 28r. The signature was later erased with ink, but it is still partially legible to the naked eye, and even more so with the help of infrared (Plate 21).

Plate 21 – Biblioteca Apostolica Vaticana, Pagès 1, f. 38r (detail).

In order to understand what is written here, we have to break up the image still visible through the erasure in order to identify letters and words.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 105

27-Sep-21 8:08:20 PM

106

chapter iv

The first group on the left is made up of the following letters arranged as a sort of “wheel,” some upright and others upside-down from the viewer’s perspective

ÁA

A L

LC

Figure 1.

By rearranging the letters into a line, we obtain:

AL Á ALC (Plate 22)

Plate 22 – Biblioteca Apostolica Vaticana, Pagès 1, f. 38r (detail).

We propose to read the letters that we have just deciphered in the following way:

Al(binus) vel Alc(uinus) On the right side of this group of letters we have two Tironian notes, one of which is upside-down like some of the letters in the “wheel” just analyzed (Plate 23). The two notes, placed on top of one another and not upside-down, mean: legi/ legit, delevi/delevit. Plate 23 – Biblioteca Apostolica Vaticana, Pagès 1, f. 38r (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 106

27-Sep-21 8:08:24 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

107

The note that amounts to a “legit” has many analogues in repositories of Tironian notes and does not present any difficulty.89 The note indicating “delevi” is similar to several notes in the margins of Carolingian manuscripts (Plate 24).

Plate 24 – a) b) Reims 375, ff. 125r, 291r (details); c) Biblioteca Apostolica Vaticana, Pagès 1, f. 39r (detail); d) A. Capelli, Dizionario delle abbreviature latine ed italiane, Milano 2011, p. 506, 2 col.

If we put the letters arranged as a “wheel” and the two Tironian notes together, we obtain the following text:

Al(binus) vel Alc(uinus) legi/ legit, delevi/delevit

Plate 25 – Reconstruction of the Albinus’signature.

89 Schmitz, Commentarii notarum tironianum, plates 25, 60 and 64; Costamagna – Baroni – Zagni, Notae tironianae, p. 8, col. 1.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 107

27-Sep-21 8:08:25 PM

108

chapter iv

In order to confirm the proposed interpretation, let us examine three more letters, slightly above and to the left side of the group of letters and Tironian notes just analyzed.90

Fig. 2.

They were written by the same hand that drew the “wheel-shaped” letters and are arranged according to the same compositional principle, since one letter is upside down with respect to the other two letters (Plate 26). The letters are the following:

Plate 26 – Biblioteca Apostolica Vaticana, Pagès 1, f. 38r (detail).

In our opinion, this group of letters should also be interpreted as an abbreviation for

Alc(uinus) The upside-down arrangement of the letters of a personal name may seem odd and unusual to a modern reader, but it was not at all odd for a 90 There are other letters written some time later on the same margin, below and on the right side of those just taken into account. In the upright position, we seem to catch a glimpse of an “E” and an “A,” which were written above another note in an upside-down position with respect to these two letters, which can perhaps be read as “Max….”

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 108

27-Sep-21 8:08:32 PM

109

Arbeo, Leidrat, Alcuin, and the codex pagesianus



reader of the Early Middle Ages, used to this type of overturning of letters in Roman epigraphs from Late Antiquity,91 and above all, to the upside-down arrangement of the letters of names in the epigraphs of churches in Medieval Ireland and England, an arrangement still visible in the early medieval inscription at St. Caimin Church and St. Graveyardare (Clare county),92 as well as in some medieval inscriptions at Galles93 and Norfolk.94 IX. Similar notes Many other abbreviations confirm the proposed interpretation. They are sometimes written in very small characters, scattered in the margins of the manuscript, almost always erased or revised, combining in various ways the same letters as those in the first abbreviation, often jumbling them. They repeat the word “Alcuin,”95 which at one point, on f. 5v, is even written in full, weaving the letters between the lines and arranging them diagonally from the bottom of the page to the top:

Fig. 3. 91 Cappelli,

Dizionario di abbreviature, p. XV. The History and Antiquities, pp. 93-174, in particular p. 153, pl. XX; id., The “Druuides” Inscription, pp. 227-238; id., Corpus Inscriptionum Insularum Celticarum, II, pl. XXXVIII; MacManus, A guide to Ogam, p. 61; Petrie, Christian Inscriptions, pp. 40-41. 93 Gresham, Medieval stone carving, p. 106 and 146. 94 M. Champion, Medieval Graffiti Inscriptions found in All Saints Church, Litcham, in Norfolk Archaeology 46 (2011), pp. 199-208, in particular p. 206, pl. 9. See also, the analysis done by the same author of graffiti in the Norfolk cathedral with names written upside-down, a parallel phenomenon and very similar to the one with that of writing names upside-down. (id., Ill wishing on the walls: late medieval graffiti curses from Norwich cathedral, in Norfolk Archaeology 46 [2014], pp. 61-66). 95 Instances can be found nearly anywhere, for example, on ff. 5v, letters woven between the lines of the text; 11r, lower margin; 48r, right margin; 58v, left margin; 66v, lower margin; 88v right margin; 95r, right margin; 105v, lower margin; 112v, lower margin 92 Macalister,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 109

27-Sep-21 8:08:32 PM

110

chapter iv

The difficulty involved in reading some of these annotations makes it difficult to compare the hand that wrote them with the hand that wrote the initials on f. 39r and the corrections to Boethius's commentary. Some of them, however, are still completely legible and were certainly written by the same hand that wrote the first abbreviation with the same type of pale brown ink. We should note that the mixture used for the ink was so diluted that the color has often vanished, and one can see what had been traced only with the aid of ultraviolet light. In some cases, reading in this way reveals, surprisingly, small drawings, one of which, on f. 73v, is a calligram made using the letters of the name “Alcuinus”

Fig. 4.

Among the various abbreviations that we find in the margins of the Pagesianus, some are constructed in a unique way. First of all, we must consider what we find on f. 39r, a little higher and to the left of the group of Tyronian letters and notes that we have already analyzed and written by the same hand (Plate 27). This group of letters is arranged in a particular way which recalls the so-called versus intexti. It can, in fact, be read vertically, from the bottom to the top, as is normal in acrostics,96 but also, in part, horizontally. In the first case, the vertical letters, read acrostically, are “A, L, C, S” accompanied by the Tironian note for “Legi” composed of “L + I” that we have already seen. In the second case, the words, read horizontally, are “L, U + I.” In both cases, we have an abbreviation that expands in the same way: the 96 On

this topic, see Pozzi, La parola.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 110

27-Sep-21 8:08:32 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

111

Plate 27 – Biblioteca Apostolica Vaticana, Pagès 1, f. 39r. The upper part of the table is taken from a photo taken in natural lighting; the lower part has been obtained by greatly increasing the contrast of a photo taken with ultraviolet rays.

first is an abbreviation by contraction formed by the letters “A, L, C, S” that reads: “ALC (uinu) S; the second is a special abbreviation, what we might call a “crasis-abbreviation,”97 formed by the letters “L, U + I,” extracted from the center of the name (a) L (c) UI (nus). A similar aesthetic, based on acrostics to be read vertically as well as horizontally, is often found in Christian epigraphs, in which, under certain names written vertically, some letters are highlighted, adding a symbolic meaning beyond that of the given name. For example, in a well-known epigraph, in the name of a woman, “Urses,” written “Vrses,” the letters “V” and “R” are enlarged, slightly detached from the others, and united by the symbol of Christ, indicating that “Christ wishes for the deceased life (V) and resurrection (R).”98 In the same spirit, the monogram of a name could be shaped in such a way that it displays a “key” meaning, such as the epigraph of Leonia on the so-called “g wall” under the papal altar of the 97 The abbreviation through crasis is obtained by isolating certain letters inside the word. The typical case is that of “ergo,” which, in the Middle Ages, and above all in the scholastic period, is abbreviated by writing a “g” with a small “o” overhead. This type of abbreviation is one of the more rare ones, but it had its own distribution. (cfr. Cappelli, Dizionario, p. 429. 98 Guarducci, Scritti scelti, p. 498.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 111

27-Sep-21 8:08:35 PM

112

chapter iv

Vatican basilica, where the “L” of the name is transformed in such a way as to contain the initials “P + E,” which indicate Saint Peter.99 This practice opens up a wide range of possibilities for “key-reading,” sometimes very complex, as in the case of the monogram for “Tire” or “Tirio” or “Tirannio,” which contains the person’s name but is at the same time an equivalent of Christ or of a Christian, since the “T” and the “Y” are the symbol of the cross, and a monogram that is formed by a “T” surmounted by a “Y” hides the cross twice.100 This method of manipulating words and combining the letters of which they are composed in different ways was familiar to Alcuin. A true heir of the best of the Irish and English traditions, the adviser of Charlemagne had a real passion for such verbal games as the “riddles” that we often find in poems, in which the answer is expected to be a word and its anagram,101 sometimes complete and sometimes partial (the so-called logogriph102). The words that make up the answer to the riddle can be read in many different ways, just as the name of Alcuin could be read vertically as well as horizontally. This taste for the anagram combined with the acrostic helps us understand another abbreviation that can be expanded in the same way. At f. 48r, on the lower right-hand edge, the letters “A, C, L, U, I,” although partially erased, can be seen running vertically from the bottom upwards.

Fig. 5. 99 Ibid.,

pp. 441-442. plurality of symbolic interpretations of similar texts is often associated with polyvalent images. In an epigraph from the cemetery of Saint Callistus, the Cross is accompanied by a depiction of Ulysses, a character who was considered to be a symbol of Christ. (Gardthausen, Das alte Monogram, p. 148 and nt. 337). 101 For a list of these compositions and an adequate bibliography, we limit ourselves to recommending the following work to the reader: Clavis des auteurs latins, II, pp. 16-18. 102 On this topic, see Pozzi, Poesia per gioco; Bartezzaghi, Incontri con la Sfinge; id., Scrittori giocatori. 100 The

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 112

27-Sep-21 8:08:35 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



113

The letters are an anagram of the name Alcuin: “A, C, L, U, I” = ­ LCUI(nus). We cannot state with certainty that the hand that drew these A letters is the same one that wrote the notes on f. 39r, because the letters, covered with ink in order to erase them, are barely visible and do not admit of in-depth analysis. Indicating someone’s name in the form of an anagram is not strange, of course. We have numerous examples of this practice in the Carolingian age,103 one of which, that of the scribe Adaboldus who signs an abbreviation of his name in an anagram, is very close to Alcuin both spatially and temporally (see Plate 34 below). We read under very different conditions a strange figure in the lower left margin of f. 76v and in the lower right margin of f. 77r of the Page­ sianus. We cannot say that it was traced by the same hand that wrote the first two abbreviations, but it was certainly written with the same ink and with the same thin pen as that of the two abbreviations on f. 39r. The figure looks like this:  

Fig. 6.

This figure could be a monogram formed by the letters “A, L, C”: In the margins of the Pagesianus we also find another type of monogram, one which occurs twice, on f. 85r and f. 95r. It was written by a hand that uses the same ink as the corrections on f. 39r and is formed by the letters “A, L,” arranged in the following way:

Fig. 7.

103 Bischoff, Mittelalterliche, 3, pp. 123; id., Die südostdeutschen Schreibschulen, pp. 190, 230-231.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 113

27-Sep-21 8:08:36 PM

114

chapter iv

X. Abbreviations and ambigrams From a purely textual point of view, the first abbreviation with which we are dealing fits perfectly with what we know about Alcuin. In fact, we know that it did not matter to him whether he referred to himself with a single name or with a combination of his own name and a nickname, e.g. “Flaccus Albinus” or “Publius Alcuinus.”104 On the other hand, his contemporaries typically referred to him by using either his two names simultaneously or his name and his nickname. The example of Einhard is representative; in his Vita Caroli, he gives the name “Albinum cognomento Alcuinum,”105 which corresponds exactly to our abbreviation “Albinus vel Alcuinus,” because “cognomento” means “called” or “named” and “vel” in our abbreviation is equivalent to “alias,” the usual way of indicating the name or nickname by which an individual was known other than his or her official name. Over the centuries, Alcuin was very often referred to in this way, using formulae like “Albinus vel Alcuinus,”106 “Alcwinus … dictus Albinus”107 or “Alcuinus … vocatus Albinus”108 or “Alcuinus sive Albinus”109 or “Alcuinus, seu Albinus (Flaccus).”110 From the point of view of writing and aesthetics, it is necessary to dwell on a series of complex questions, keeping in mind the fact that, in every case, the abbreviation “Al” in Rustic

Plate 28a – a) Zürich Zentralbibliothek, C 80, f. 83v (detail); b) Paris BNF Lat. 2974, f. 50r (detail); Biblioteca Apostolica Vaticana, Pagès 1, f. 39r (detail). 104 For

example, Epistolae, XXVII, XLII, XLIII, XLIV, LXXVI, C, and CVII. Vita Karoli, p. 25; M. Tischler, M. Einharts Vita Karoli. 106 Raniero Giordani da Pisa († 1348), Pantheologia, pp. 263, 646; Stengel-Reihing, Nova S. Scriptvrae, p. 263; Cornelii a Lapide, Commentarius, p. 8. 107 London, British Library, Arundel 218 C, f. 1v (XIII Ct.). 108 Ibid., f. 2v (XIII Ct.). 109 Trithemius, Liber de scriptoribus, f. 41r. 110 Alcuinus seu Albinus opera de novo collecta, emendata et aucta cura et studio F. Frobenii, Ratisbona 1777. 105 Eginardus,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 114

27-Sep-21 8:08:36 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

115

Capitals for “Alcuinus” is precisely the one used in authoritative manuscripts of his works in the ninth century111 (Plate 28a). In the first abbreviation on f. 39r, the letters seem to be placed in an almost circular shape and are partially overturned112: a similar arrangement 111 In certain copies of the Alcuinian “dialogues” from the ninth century, the form “Al” is a frequent indicator of the author’s identity, as, for example, in Paris, BnF, lat. 2974, f. 50r or in Zürich C 80, c 33v. In the majority of cases, the form “A” predominates as a stand-in for “Alcuinus,” as, for example, in Bern, BB 234; Munich, Bayerische Staatsbibliothek, clm 6407; Orléans, BM, 273; Paris, BnF, lat. 7581; Sankt Gallen, SB, 272 and 273. 112 The inversion of a letter or word occurs occasionally or systematically in many writings from very different time periods. Typically there is a reason which explains this sort of phenomenon. For example, in Roman epigraphs we usually find a woman’s name or the initial letter of her name upside-down with respect to those of men. In the scrolls of the Exultet, written in the Beneventan script of southern Italy, the text which the assembly present at the ceremony needs to read is written upside-down with respect to that of the priest at the altar, so that the unrolled scroll might present itself in reverse, physically speaking, to the one who is performing the ceremony. In any event, the most common reason for the inversion of letters is surely the magical use of the words in the context of what Mauss calls the “general theory of magic.” Inverting the name of an individual means, in every civilization, subverting his existence in order to cause the “inversion” of his life (Faraone – Kopp, Inversion, Adversion and Perversion, pp. 381-398). He thus inverts the name of the man for whom he wants everything to go “upside-down.” Numerous examples of this special form of enchantment can be found in every part of the world and in every age, in particular, to stay with the region with which we are dealing, in the Mediterranean basin, in the Near East, and in ancient, late-antique, and medieval North Africa, in the so-called tabulae defixionis, in magical papyri, and in many other texts written in Greek, Latin, Coptic, Syriac, Hebrew, and Arabic. In these same geo-linguistic regions there are also many accounts of attempts to counteract and remove such curses by resorting to apotropaic formulae in which we find the same sort of inversion of the names and letters of the evil that one wants to avoid. Whoever does not want his life to be “inverted” can “invert the inversion” and overturn, so to speak, the curse by inverting the very letters with which the curse was written. European and Middle Eastern folklore is full of examples of “inverting the inversion.” A significant example of this kind of magical act is the habit of writing the name of a dead person on the tombstone while inverting the letters which spell it out. This apotropaic custom, widespread in medieval Ireland, Scotland, England, and Wales, is of particular interest to us for the purposes of the narrative which we are creating. In early medieval tomb epigraphs at St. Caimin’s Church and St. Graveyardare (Clare county), as well as in some medieval Welsh inscriptions and graffiti on the churches of Norfolk, we find the letters of names inverted systematically in order to remove all evil from those designated by the names, as well as to allow the dead to pass over to the Great Beyond while preventing them from returning to the world of the living. A recurring phenomenon in the context of all of these direct and indirect manifestations of “magical thinking” is the use of a particular layout for writing curses and the formulae of spells: a square or circle, within which the writing follows the figure’s perimeter and appears partially upside-down to the reader’s eyes as a result. This aesthetic structure of writing is nothing other than one of the many manifestations of the so-called “magic square” or “magic circle” of which we have many examples in the context of witchcraft and magic in every age. The practice appears already in the Roman tabulae defixionis (Faraone – Kopp, Inversion, Adversion and Perversion, p. 384). We find it also in circular and square-shaped amulets from antiquity which are then imitated in the early Middle Ages. This method of writing

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 115

27-Sep-21 8:08:36 PM

116

chapter iv

of letters appears in ancient times in monograms on Greek coins of various time periods that come from Asia Minor or Thrace, in Latin inscriptions of Italian origin from the early imperial period, and in Roman coins from the time of the Emperor Hadrian113 (Plate 28a, b).

Plate 28b – Ancient monograms: a) “mi, epsilon, alpha, sigma” for MES[EMBRI]A); CIL, XIII, 10009, 30: “A,N,T, I” in latin); c) “alfa, pi, lamba, omicron” for APOL[LONIA]). V. Gardhtausen, Das alte Monogramm, Leipzig 1924, 2, n. 69; 4, n. 143, 2, n. 68).

This sort of writing technique was often employed in the Carolingian age and later. Typical examples are the monograms of rulers such as Louis the Pious or Eudes I114 (Plate 28c). Following the same rules, you can build monograms in the form of an ambigram.115 Ambigrams are a sort of optical illusion, common in Late Antiquity and the Middle Ages, in which a word can be read in the same ritual formulae or the names of people in a circle, at times indicating their role, becomes so widespread and common that it loses its magical significance and can be used any time one wants to solemnly the name of a person and his office. We find this sort of remembrance in the form of a rota in signatures and seals on public documents issued by the chancelleries of European monarchs such as the Merovingian kings and those of Castille, where they are intrinsically associated with very different manifestations of the signum recognitionis of the authority issuing the document. In them, as in the case of the amulets, the text of the fixed formula, which records the name and office of the one who sanctions the document, often runs around the entire circle and thus appears partially upside-down to the eyes of the reader. 113 Ibid., pp. 25, 35. 114 Diplomata Karolinorum, II, plates 47, 79; VIII, plate 4. 115 Hofstadter, Ambigrammi.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 116

27-Sep-21 8:08:37 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

117

Plate 28c – The monogram of Zeleia formed by four letters: “zeta, epsilon, lambda, epsilon”. The “lamba” and the second “epsilon” are upside-down with respect to the “zeta” and to the first “espilon”(Copper coin dating from 350-300 Af. Troade, Zeleia: Sylloge Nummo­ rum Graecorum, edited by H. von Aulock, Berlin 1957-1968, 7663).

Plate 28d – Monogram of Louis the Pious and of Eudes I.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 117

27-Sep-21 8:08:37 PM

Plate 29a – A “perfect” ambigram formed by two identical signs, intertwined in such a way that one always reads the same letter, even if the page is turned upside-down. Mozarabic Antiphonary (tenth century), Léon, Archivio Catedralico n. 8, f. 5r.

Plate 29b – An “imperfect” ambigram: the group of letters “T+A+V” in the inscription of Saint Clement of Taull of 1123. The ambigram, consisting of two letters which are apparently the same, an “A” and a Visigothic “V” with a horizontal stroke, apparently “perfect” because it seems to be formed from two equal parts, is nevertheless “imperfect” because a small section has been added to the top of the “A,” indicating a “T” in connection with the “A.”.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 118

27-Sep-21 8:08:37 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



119

way if it is turned upside-down. Likewise, a monogram constructed as an ambigram can be read just as well even if it is turned over. A typical example of the latter is the monogram of the bishop Otto of Freising († 1158) which is always the same even if it is turned upside down.116

Fig. 10.

The abbreviation which we found in Pagès 1 in the margins of f. 39r was conceived in the same way; it is an “imperfect” ambigram. This perfectimperfect distinction is applicable to many medieval ambigrams, which in some cases are perfectly reversible and in other cases are only imperfectly or partially reversible, but reversible nonetheless (Plates 29a-b). The abbreviation on f. 39r of the Pagesianus can be considered an “imperfect” ambigram because, even if there are two identical letters that can be read by turning the text upside-down, there are also two small, non-reversible letters to the viewer’s left (“LC”), deliberately made smaller so that the viewer does not notice and the” optical illusion “of having an ambigram in front of him is preserved:

Fig. 11.

In any case, whether perfect or imperfect, the ambigram reads as a “rota,” most likely corresponding to a real “rota” used by those who sign a signum recognitionis in documents. 116 Leist,

Urkundenlehre, plate III.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 119

27-Sep-21 8:08:37 PM

120

chapter iv

We can compare this wheel with the initials of Adabaldus,117 a famous copyist active in Tours under the Alcuin’s successor, the abbot Fridegiso (804-834) who signs it (plate 29c) on the sheet kept in Paris, BnF, Nouv. Acq. Lat. 405:

Plate 29c – Paris, BnF, Nouv. Acq. Lat. 405 (a leaf without a number). The signature of Adabaldus.

The abbreviation of the name Adalbaldus consists of the letters “A, L, D, B” that form the anagrammatic name of the scribe (the correct progression of the letters should have been “A, D, L, B,” since the name which that same amanuensis writes in full is “Adabaldus”118). We should also emphasize that the letters “turn,” so to speak, in a circle with respect to the reader, but there is one anomaly: the “D” that follows the “L” should be reversed with respect to the viewer if we want to keep the reading order from left to 117 Delisle,

Mèmoire de l’Ecole calligraphique de Tours, p. 20, pl. I-IV; Desnoyés – De­ Note sur un monogramme, pp. 376-381; Rand, A Survey, pp. 56-58; Catalogue des ma­ nus­crits en écriture latine, IV, 1, p. 85. 118 In another case, the same copyist signs “Aldabaldus” (Delisle, Mémoire, p. 20): but, in this case, things are not the same, and the only possible explanation of the inversion of letters in the monogram is that we are dealing with an anagram of his name. lisle,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 120

27-Sep-21 8:08:37 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



121

right between the first and second letters. Instead, it is straight from the viewer’s perspective, creating an alienating effect for one attempting to read it normally, just as the reversal of letters creates a sense of alienation from normal ways of reading. XI. Anagrams and acrostics Analyzing the second abbreviation on f. 39r, we must observe that the letters are written vertically and can be read from the bottom upwards but also, in part, horizontally. This way of reading words is typical of extremely elaborate compositions such as carmina figurata and acrostics.119 This abbreviation in the Pagesianus is similar to ones we find in two witnesses from the last quarter of the eighth century and the beginning of the ninth, in which we find the name Alcuin written once vertically and once vertically and horizontally with a visual effect very similar to that of the Vatican manuscript. The first abbreviation is in a poem dedicated to Charlemagne by Alcuin, which begins “Magna quidem pavido.” As seen in a ninth-century copy of the poem in Bern, Burgerbibliothek, 212, the name “Albinus” appears in the center of the composition, running vertically. The second one is in the Bamberg Bible (Bamberg, Staatsbibliothek, Bibl 1, f 5v), copied, according to Rand, at the beginning of the ninth century in Tours, and, according to other scholars, around 834 in Marmoutier, an abbey supervised by Alcuin together with that of Saint Martin of Tours.120 In this manuscript is the only portrait of Alcuin that can be considered coeval or almost contemporary. The portrait’s caption — unique compared to those of all the other portraits in the manuscript — consists of two groups of letters, one that reads vertically and consists of “A, L” (which, as we said, was a Carolingian abbreviation for “Alcuinus”) and the other horizontally, in which we find “C, U, I, N, U, S”121 and then “ABBA.” The two groups of letters form the 119 Pozzi,

La parola. this famous codex, we refer, for the sake of brevity, to Suckale-Redlefsen, Die Handschriften des 8. bis 11. der Staatsbibliothek Bamberg. I, pp. 26-30, with a full bibliography. 121 It is also possible that in the didascalia of the Bamberg Bible one of the logogriphs or partial anagrams that Alcuin liked so much is hidden. The letters “cuinus” of the name “Alcuinus,” which can only be read horizontally, could be an anagram of the word “unicus.” The word “unicus” could refer to an idea repeatedly affirmed by Alcuin in his polemics against the Antiadoptionists (“Unicus igitur est Christus Filius Dominus noster, non velut coniunctione qualibet et unitate dignitatis et auctoritate hominis habentis ad Deum, quem tu soles coniunctum Deo, sive adoptatum vocitare, divinitatem quoque gestare, et, nescio quo insolito verbo usus, divinitatem liniare dicis.” Alcuinus, Contra Felicem, V, 6, in PL 101, col. 214 D); or it could simply highlight the fact that he was the “unicus abba” of Tours and Marmoutier. 120 On

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 121

27-Sep-21 8:08:38 PM

122

chapter iv

name “ALCUINUS ABBA” (Plate 36). The similarity between the structure of this abbreviation and the one in the second abbreviation of the Page­ sianus is evident: the method of reading is simultaneously vertical and horizontal, the same as the so-called versus intexti, even if it is not a carmen fig­ uratum. This construction of a caption such that it can be read in two ways is certainly not common. It is, however, occasionally found in manuscripts of the Carolingian age, such as the Godescalco Gospels produced at the court of Charlemagne,122 in which, at f. 4r, the words “In illo tempore” can be read vertically while the following words — the opening words of the Gospel of Matthew — are read horizontally. Another similar example is the incipit reported by Bischoff123 on f. 1r of Paris, BnF, lat. 1868, in which the letters are read simultaneously vertically and horizontally. One example is the word “Liber”: the first two letters “L, I” from bottom to top, the next two letters “B, E” from top to bottom, and the last letter, “R” horizontally:

Fig. 12.

The abbreviations in the Pagesianus and in the Bamberg Bible already discussed are constructed in a similar way (Plate 30).

Plate 30 – a) The name “Alcuinus” in acrostic in the versus intexti of the poem “Magna quidem pavido” Bern, Bürgerbibliothek 212, f. 20r; b) The name “Alcuinus” written in the form of an acrostic abbreviation in Bamberg, Staatsbiliothek, Bibl. 1, f. 5v; c) Biblioteca Apostolica Vaticana Pagès 1, f. 39r (excerpted). 122 Das Godescalc Evangelistar. See also Bischoff, Katalog, p. 44; Denoël, Entre imitation et invention, pp. 89-120; ead., La vallée de la Loire: de Saint-Martin, pp. 151-165. 123 Bischoff, Mittelalterliche Studien, 3, p. 121.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 122

27-Sep-21 8:08:41 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



123

The unique construction of the letters of Alcuin’s name in the figurative poem and in the Bamberg Bible confirms what we said earlier: we are in the realm of the acrostic, a fact proved by the ability to read a text written horizontally from top to bottom and sometimes even from bottom to top. Confirmations Manuscripts dating from the last years of Alcuin’s life or immediately following his death confirm what has been said so far. A significant example is a manuscript written in Tours while Alcuin was still alive. In it we find both an anagram of “Alcuinus” and an acrostic — which reads both vertically and horizontally — containing the anagrammed letters of Alcuin’s name, an acrostic very similar to the one in the second manuscript. The examples which we are discussing are on the first and last leaves of Paris, BnF, lat. 1572, a manuscript of the Acts of the Council of Ephesus written in Tours between the end of the eighth and the beginning of the ninth century,124 a manuscript which certainly bears traces of the cultural work carried out by Alcuin and his collaborators, and which, according to Bischoff, would have been annotated by Alcuin himself, a hypothesis not accepted by some scholars. On f. 225v, at the end of the manuscript in whose margins would be this presumed Alcuinian autograph, a hand contemporaneous with that of the text has, in fact, written the letters “A, C, L” in a large format, letters which are only partially legible today.

Plate 31a – Paris, BnF, lat. 1572, f. 1r, lower right margin. 124 Rand,

A Survey, pp. 87-88.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 123

27-Sep-21 8:08:45 PM

124

chapter iv

In the same Paris manuscript, at f. 1r, a hand contemporaneous with that of the text has outlined a sort of calligram forming a kind of pyramid. The pyramid is composed of five letters that are repeated twice (A, C, L, E, F) and that form different words or different abbreviations of words in partial anagrams or logogriphs, legible both vertically and horizontally. This half-erased pyramid slants along the sheet like the abbreviation in the Pagesianus and is constructed according to the technique of versus intexti (Plate 31a, b).

Plate 31b – Paris, BnF, lat. 1572, f. 1r, lower right margin.

This strange figure composed of repeating letters can be deciphered by combining the method of reading a partial anagram with an expansion of its abbreviations in the following way (Plate 31c):

Plate 31c – Paris, BnF, lat. 1572, f. 1r, lower right margin.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 124

27-Sep-21 8:08:54 PM

Arbeo, Leidrat, Alcuin, and the codex pagesianus



125

The reader who is not familiar with these kinds of verbal games, which were widespread in Late Antiquity and the early Middle Ages, could find himself surprised and a bit skeptical. To change his mind it will be enough to consult the poems of Alcuin and dwell on the ones that, in tune with the verses of the best Insular poets like Aldhelm or Tatwin, propose puzzles that are solved with total or partial anagrams of words and names.125 Let it suffice to mention the carmen that begins with the phrase “Causa nocis fueram,” a riddle whose answer consists of an anagram that generates two words comprising the same letters arranged differently: “malum” and “mulam.” Likewise, possible solutions to the next poem, which begins “Sex mihi litterulae sunt” are four words derived from a partial anagram or logogriph of a group of letters, namely “Virtus,” “tus,” “vir,” and “virus.” The response to the poem that begins: “Vir totus canutus” consists of a partial anagram of the same set of letters: “Sanus,” “anus,” and “sus.” Continuing to read, one reaches the vertex of virtuosity, a height represented by the poem that begins “Presbyter integro celebravit missam,” which contains a series of partial anagrams of identical letters: “Magnus,” “agnus,” “manus,” “Magus,” “mus,” and “anus.” Alcuin, who called himself Publius Alcuinus in honor of that famous author of versus intexti Publius Porphyrius Optatianus,126 wrote total anagrams, logogriphs, and carmina figurata127 (two adressed to Charlemagne). He directly inspired his disciple Rhabanus Maurus,128 with whom he is depicted in the most ancient and authoritative manuscripts of his collection of figurative poems known as the Liber de laudibus Sanctae Crucis.129 Other authors at court of Charles, such as Teodulfus, Paolinus of Aquileia, and Joseph the Scot, wrote poetry in the same genre using versus intexti.130 In this cultural, aesthetic, and literary atmosphere, both the anagram and the transcription of Alcuin’s name such that it can be read vertically and horizontally are quite natural. It is, therefore, not at all strange to find the practice recalled in this way in the second manuscript, Pagès 1. There are at least two other confirmations of our interpretation in the manuscript Médiathèque du Grand Troyes (formerly Bibliothèque Municipale di Troyes), Fonds ancien 1742. According to all scholars, the manu125 Alcuinus,

Carmina, n. 63, pp. 281-282. Optatiani Porfyrii Carmina, ed. G. P. Polara. Cfr, Deroux, The Carmina of Publilius Optatianus Porphyrius, pp. 447-466. 127 Chazelle, The Crucified God. 128 Perrin, La poésie de cour carolingienne, pp. 333-335. 129 Spilling, Opus Magnentii Hrabani; Ferrari, Il “Liber sanctae crucis.” See also Gu­ glielm ­ etti, Hrabanus Maurus. 130 Ernst, Carmen figuratum., pp. 158-202. 126 Publilii

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 125

27-Sep-21 8:08:54 PM

126

chapter iv

script was produced in Tours during Alcuin’s life.131 In its margins we find frequently recurring abbreviations, erased but still partially legible, formed from the letters “A, C, L,” “A, L,” and “A, L, U, I,” abbreviations quite similar to those found in the Pagesianus. Among these, two abbreviations in particular draw our attention. The first one is constructed exactly like the one in Pagès 1 at f. 48r (Plate 32a).

Plate 32a – Troyes, Médiathèque du Grand Troyes, Fonds ancien 1742, f. 67v.

The second abbreviation, at f. 84r, is similar to the “rota” of f. 39r in the Pagesianus (Plate 32b).

Plate 32b – Troyes, Médiathèque du Grand Troyes, Fonds ancien 1742, f. 84r. 131 Bischoff, Katalog, III, p. 386. On the manuscripts of Alcuin, we refer, for the sake of brevity, to Ganz, Handschriften, pp. 185-194.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 126

27-Sep-21 8:09:03 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

127

As in the case of the “rota” in the Pagesianus, we are reminded of the letters in the abbreviation of two names, joined by “vel,” by which a person is known. Here they can be seen in the form “ACLI vel AL,” obviously corresponding to “ALC (u) I (nus) vel Al (binus).” Last but not least, on the same page of the Troyes manuscript we find Alcuin’s name written vertically and in full by the same hand that drew the initials parallel to it (Plate 33).

Plate 33 – Troyes, Médiathèque du Grand Troyes, Fonds ancien 1742, f. 84r. On the left is a detail of the original sheet, which shows the name “Alcuinus” written vertically and slantwise so that it reads from bottom to top. On the right is the same name as it appears in the original, next to a reconstruction of how it would look if we turned it over to read from top to bottom.

XII. Alcuin’s hand? Is this signature, its letters upright as well as upside-down, the autograph of Alcuin? It seems likely. Or, at least, likely that the notes were written under Alcuin’s eyes, if not by him, by a secretary. In any case, there are no reliable examples of Alcuin’s autographs against which to compare these notes. Mayer and Bischoff have suggested the only two possible examples, both dating from near the end of the abbot of Tour’s life, but these suggestions have recently been challenged by more than one scholar132 and 132 Jullien, Alcuin et l’Italie, pp. 393-406, in particular pp. 398-399: “On ne saurait pour­ tant déduire que l’une des deux mains du IXe siècle soit celle d’Alcuin plutôt que de l’un de ses

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 127

27-Sep-21 8:09:08 PM

128

chapter iv

are not very useful. For this reason, one should recall an observation made by Michael Lapidge, emphasizing the difficulty of identifying autographs from the learned Englishman’s last year of life, during his residence in Tours, because of his health. Lapidge reached the conclusion that, at this stage, “Alcuin was old and arthritic with a failing eyesight (as he reminds Charlemagne in a letter of 799) and although his scholarship continued unabated, he must have depended increasingly on amanuenses.”133 Nonetheless, despite these limitations of advanced age which beset him after his appointment as abbot of Tours, it is not impossible to imagine that Alcuin personally annotated his manuscripts before going to Tours with handwriting that looked like Caroline script while retaining traces of Insular writing. He may have been assisted in his work by a secretary, as we see in the corrections to the Pagesianus which would have completed the work of revision begun by the master. The hypothesis is not impossible and can be maintained on the grounds of some other observations. First of all, it is necessary to point out that the hand we are considering cannot be dated to the time of his transfer to Tours, after 799, since, as we have said many times, the Pagesianus manuscript was in Lyon with Leidrat from 798 onwards. Moreover, as Holtz rightly pointed out, the Pagesianus was a part of Leidrat’s “bibliothèque personnel”134 and was not meant to circulate during the years of his episcopate. But if this is true, we cannot attribute the corrections to Boethius’s Commentary to the hand of some erudite man residing in Lyon, capable of correcting a manuscript that Leidrat had just brought into a city that did not know its content, one that would have been kept carefully among his personal books. Once it is established, therefore, that these annotations were written collaborateurs. Elles diffèrent d’ailleurs de la main identifiée par Bernhard Bischoff comme celle d’Alcuin en marge du f. 79r de Paris, Bibliothèque nationale de France, lat. 1572 (Actes du 1er Concile d’Éphèse copiés à Tours durant la 2e moitié du VIIIe siècle, et utilisés par Alcuin dans ses écrits contre l’hérésie de Félix d’Urgel) et aux ff. 20v-22v de Gotha, Forschungs- und Landesbibliothek, Memb. I 75 (Carmina rythmica ajoutés vers 800): hypothèse à laquelle il est difficile d’adhérer car, même si Alcuin a utilisé le premier manuscrit, les deux seules lettres «d» et «S» (pour dimitte et scribe) signalant les passages à recopier sont un bien maigre indice pour conclure qu’il s’agit de sa main; elles ont pu être inscrites par l’un de ses collaborateurs. Quant au passage du manuscrit Gotha qui serait de la main d’Alcuin, il s’agit de poèmes dont les nombreuses fautes textuelles démentent qu’il puisse en être l’auteur; de plus, le tracé du «d» ne ressemble pas à celui de Paris 1572.” 133 Lapidge, Autographs of Insular Latin Authors, p. 121. 134 Holtz, Leidrat, pp. 322-323. The general reconstruction of Leidrat’s writing and cultural activity made by Holtz is, as always, of high scientific value. We should rememember that, as is natural for a serious and well-read scholar, Holtz considers as authoritative and beyond dispute the opinion of the majority of scholars on the dating and origins of the Pagesianus, reporting it as something certain.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 128

27-Sep-21 8:09:08 PM



Arbeo, Leidrat, Alcuin, and the codex pagesianus

129

before Leidrat’s appointment as bishop, we have another remark to make: who could ever have been so bold as to amend the manuscript so jealously cherished by Leidrat, and in whom did this somewhat impudent urge arise of correcting it while in possession of an incomplete antigraph? This point demands careful consideration. Theoretically, any well-read personality at the court of Charlemagne might have done it, but, practically speaking, this is very, very difficult. Nobody, in fact, had the culture and authority necessary for the task. It presupposed so much knowledge of Boethius’ text, to the point of allowing oneself the luxury of amending it and occupying a position hierarchically above that of Leidrat, a man personally protected by Alcuin and called filius noster by him. To be sure, in the Pagesianus there were quite obvious mistakes (as, for example, peripatica instead of peripatetica) and omissions of titles and passages, enough to justify the process of revision. It is just as true, however, that the corpus of logical works in Leidrat’s volume was an absolute novelty, so that most readers would be discouraged from dealing with it and, above all, from doing it with excessive impatience without having a better (whole) copy of the text at hand. On the other hand, this sort of activity seems entirely appropriate for a man like Alcuin, who in precisely those years did his very best, with all of the enthusiasm of a neophyte, to circulate these texts, even sending them to Charlemagne with verses of dedication, plundering them in order to compose little works dedicated to the ruler himself or to his friends. But there is more to the story. It does not seem to be a coincidence that the marking of some passages meant to be extracted corresponds precisely to the passages of the Boethian Commentary quoted verbatim in the Libri Carolini.135 We know that the Libri Carolini brings together all of the knowledge of the most learned representatives of the Carolingian Renaissance, whose chief architect was Alcuin. And we mentioned above that the formulations of the Creed appended to the end of the Pagesianus are quoted word-for-word in the Libri. Indeed, it could not be a coincidence that passages of this very rare Boethian commentary, discovered by and brought to the attention of an avid reader of the Pagesianus, would later be extracted from the manuscript of the Marist Fathers on the recommendation of a hand that wrote in a Caroline minuscule full of Insular elements, 135 For the sake of accuracy: f. 41r, lines 13-18 = Libri Carolini ed. Freeman, pp. 545-546 = Boethii Commentarii, I, ed. Meiser, p. 36, rr. 22-29; f. 45r, right margin, corresponding to the first two lines of the quotation of Boethius marked in the Libri Carolini, ed. Freeman, p. 545 lines 15-16 and 16-23 = Boethii Commentarii, I, ed. Meiser, p. 45, line 29 and p. 46, lines 1-6; f. 53r, lines 8-20 = Libri Carolini, ed. Freeman, p. 545 = Boethii Commentarii, I, ed. Meiser, p. 69, rr. 23-25 and p. 70, rr. 1-26).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 129

27-Sep-21 8:09:08 PM

130

chapter iv

a hand which identified itself as “AL” and “Alc,” abbreviations which could stand for Alcuinus. This process can be understood if the Pagesianus was a “working manuscript” belonging to somebody who was at the same time working on the production of other texts or who at the very least inspired those who were working on them; in short, it was the work either of Alcuin or of his alter ego. What, then, are our conclusions? We grant the reader the freedom to unravel the enigmas that we have proposed. However, as a conclusion to our study, we have to underline an obvious fact, much more important than the possibility that the annotations may really be (as indeed appears to be the case) an autograph of Alcuin. On the basis of all that we have seen in this chapter, it seems clear that it was Leidrat who brought to the Carolingian court the fresh breeze of little-known authors. The rediscovery of the old logic in the Carolingian period depends originally on Alcuin’s filius Leidrat rather than on Leidrat’s spiritual father Alcuin. This “rediscovery” began in the region of the early medieval Noricus whence came Leidrat. This region included both the Freising of Arbeo and the Salzburg of the bishop Virgilius, nicknamed “the geometer,”136 who may not have believed in the Antipodes but did read rare and unusual works, enough to earn him his nickname. It also included part of northern Italy, with its old libraries in which authors like Boethius were read and studied for ages, as happened in Bobbio, where the oldest fragment of the Roman philosopher’s De arithmetica is preserved (CLA 450) and where Gerbert of Aurillac found a copy of the same author’s De astronomia, a work irrevocably lost today.137 Other Boethian important manuscripts came from Italy, as well, such as the archetype of the oldest copy of the Consolatio (Orléans, B.M. 270) or the codex containing the logical corpus, owned by Renatus and corrected by Eusebius “vir venerabilis” in Vivarium.138 Of course, we do not at all mean to imply that the importance of a giant such as Alcuin and his work in promoting the knowledge of logic and philosophy should be denied. Volumes have been written on this topic, and it would be absurd to deny both the facts and a very rich literature, full of superb contributions. However, it would be just as absurd to refuse to acknowledge reality and deny a role in the rediscovery of the Old Logic to people other than Alcuin, people who have unfortunately been kept in the background for a long time. 136 Virgil

von Salzburg. Una pietà più profonda, pp. 703-727. 138 Cfr. supra p. 16; F. Troncarelli, Il teatro delle ombre. Scritture nascoste e immagini invisibili in codici e mosaici tardoantichi (in press). A detailed study of Eusebius is provisionally planned for publication by the Centre for the Study of Manuscript Cultures in Hamburg (FT). 137 Troncarelli,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 130

27-Sep-21 8:09:08 PM

CHAPTER V

THE DEVIL IN THE MIRROR: ALCUINIAN DRAWINGS IN THE CODEX PAGESIANUS In 790, Alcuin wrote a letter from England to his disciple Joseph, asking for clothing and money because his diplomatic mission was lasting longer than expected. He also requested something to help while away the time during his forced stay, namely “pigmenta multa de sulfure bene et coloribus ad picturas.”1 This rather unusual request did not arouse much interest,2 but it is worth further attention, because it challenges the traditional image of Alcuin as someone indifferent to the embellishment of his manuscripts.3 As suggested by Saywer, the “pigmenta” were intended “presumably to illuminate manuscripts,” a hypothesis confirmed by the presence of sulfur-based pigments. This interpretation is corroborated by the wording that occurs in Alcuin’s text, “pigmenta de sulfure bene.” In fact, the adverb “bene” is simply a contraction of the stock phrase “bene pulverizato,” often employed to refer to compounds containing pulverized sulfur.4 This corresponds exactly to what we find in one High Medieval recipe for gold dust for miniatures,5 made by mixing mercury and sulfur with gold. The mercury and sulfur were pulverized and either left unbaked or sublimated as required in order to make other colors.6 1 Alcuinus, Epistulae, n. 8, pp. 33-34. Regarding Alcuin’s diplomatic mission and the complex political situation, we limit ourselves to referring the reader to the recent volume by Dales, Alcuin, pp. 5-106 (with updated bibliography). 2 Lebecq, Marchands et navigateurs frisons, pp. 362-363; Sawyer, The Wealth, p. 66. 3 This opinion, apparently well-grounded, is quite widespread and was also recently reasserted in the stunning catalogue of the exhibition Trésors carolingiens, p. 151. 4 Hunt, Popular Medicine, p. 242, n. 
56: “Capiatis … parum de sulphure bene pul­ve­ri­ zato.” 5 Caffaro, Scrivere in oro, pp. 18; 199-201. 6 Pulverized sulfur was used to produce gold dust, but in most other instances it was used in a different form. To obtain cadmium, the sulfur was purified through sublimation (Caffaro, Scrivere in oro, p. 59) or mixed with other ingredients and boiled (ibid., p. 153); to obtain the unknown color called “antimis dedami,” it was baked together with other components (ibid., p. 65); for the artificial orpiment, sulfur and ealgar (or arsenic sulphide) were melted (Cennini, Libro dell’Arte, cap. XLVI); to obtain artificial cinnabar, sulfur was baked with other components (Caffaro, Scrivere, p. 115). In general, however, most of the

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 131

27-Sep-21 8:09:09 PM

132

chapter v

In any event, whatever we may think of Alcuin’s “pigmenta,” it is a fact that the act of requesting them reveals the future Abbot of Tour’s interest in the “pictura” and justifies an inquiry into possible manifestations of this interest. Such research requires, of course, a deep study of many meaningful pieces of evidence, and yet, as an introduction to such an ambitious project, it is worthwhile to consider the results of analyses of a single manuscript. This manuscript, according to the opinion of the most authoritative scholars, is a direct copy of a model produced in Alcuin’s circle7 or, according to the most recent research,8 was actually annotated and amended by Alcuin himself or by a secretary, working under his eyes: the manuscript Pagès 1 of the Vatican Library (once recorded as Roma, Casa dei Padri Maristi, A II 1). As we have seen in the preceding chapter, this important witness to the transmission of Aristotle in the Middle Ages has been examined and transcribed many times in the past, but a deeper, updated analysis of its script and codicological structure reveals that many of the traditional evaluations are questionable. It was common to date Pagès 1 to the time of Leidrat’s episcopate in Lyon (798-814). However, the date and attribution of the codex should be modified on the basis of new research. On the recto of the first leaf, there is a note of ownership by bishop Arbeo of Freising († 784), in whose service was the young Leidrat. This subscription, written by a hand similar to Leidrat’s and datable to the last quarter of the eighth century, allows the first quires of the manuscript to be dated prior to 784. This chronological shift is confirmed by the general appearance of the handwriting of the scribes who copied the manuscript, as well as by the similarity in appearance of several hands to the handwriting of the scribes of manuscripts originating in Freising, such as Munich, Bayerische Staatsbibliothek, clm 6309 (CLA, 1272), transcribed during Arbeo’s episcopate. In fact, besides Arbeo’s name, prefixed to the first work contained in the manuscript, we also find Leidrat’s notes of ownership at the end of each text reported in the volume. It is natural to interpret Leidrat’s notes of ownership as different ex libris, appended one by one, each at the end of each individual book, separated from the others. This opinion is confirmed by the fact that these works each have an independent fascicle numbering, revised only at a later time with an eye to making it progressive and continuous. The individual books were doubtlessly written by the same  

extracted sulfur was melted directly where it was found (ibid., p. 89). For an introduction, see Andreuccetti – Lazzareschi Cervelli (eds.), Il colore nel Medioevo: arte, simbolo, tecnica. 7 Bullough, Alcuin and the Kingdom of Heaven, pp. 37-40 in particular p. 176; Holtz, L’oeuvre grammaticale d’Alcuin, pp. 129-149, in particular p. 144, n. 47. 8 Troncarelli, L’antica fiamma, pp. 97-152.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 132

27-Sep-21 8:09:09 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

133

scribes as those who wrote the first booklet, although with different inks, as is the case with transcriptions made within a few years of one another. The texts were completed at different times, although not excessively distant from one another, and the volume is the result of a final revision and reassessment of the entire miscellany. In conclusion, we have to date the manuscript to the last quarter of the eighth century, taking into account that it is a compound of different parts written during that period of time.9 As we have already seen in the previous chapter, the manuscript conceals a fascinating mystery. From f. 39r up to f. 48r, there are frequent interlinear and marginal corrections of Boethius’ commentary on the Peri hermeneias. They were written by a hand contemporary with that of the copyist, using an ink dimmer than the one used by the scribe, essentially Caroline, but with a clear Insular influence, and they can be dated between the last quarter of the eighth century and the beginning of the ninth. The hand of this reviser of uncommon erudition, who took the liberty of correcting the text of Boethius’ commentary, an undertaking very rare at that time, signed it more than once using an abbreviation made of the following letters: “AL” and “ALC.” We find this same abbreviation on other leaves of the manuscript which contain no corrections, at times even in the form of the so-called “alphabet in images” (letters arranged to form an image), next to or within a drawing (Plates 34-35). It seems possible to interpret them as abbreviations of the name “Alcuinus,” which, among other things, we find written in full twice by the same hand, once on f. 39r in the lower margin, oblique to the one in the folio’s text, and again on f. 5v, in the very

Plate 34 – Biblioteca Apostolica Vaticana, Pagès 1, f. 113v (details). 9 The manuscript was reworked multiple times during the centuries, which caused the rearrangement of some leaves and the alteration of the original fascicle arrangement. More recently, the volume was restored and shows many traces of this intervention, surely not a refined one.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 133

27-Sep-21 8:09:09 PM

134

chapter v

middle of the page, arranged in an unusual way, inserting the letters of the name between the lines of the text and arranging them diagonally from the bottom upwards. The hypothesis of Alcuin’s direct presence in the Pagesianus which was owned by Leidrat is indeed possibile, since the future bishop of Lyon was Alcuin’s beloved disciple and remained at his side for many years — or at least stayed in touch with him — maintaining a beneficial relationship of cultural exchange. This would coincide with the observations of the most meticulous and authoritative scholars who have closely studied the issue, such as Louis Holtz, who emphasized the profound loyalty of Charlemagne’s secretary to Boethius’s commentary on the Peri hermeneias, to the extent that he claimed that “Alcuin avait sa source sous les yeux.” and that “nous avons vraisemblement conservé avec le ms. Roma, Casa dei Padri Maristi, s.n. [later on A II 1 and today Pagès 1], une copie de l’exemplaire d’Alcuin.”10

Plate 35 – Biblioteca Apostolica Vaticana, Pagès 1, f. 113v (details). 10 Holtz,

L’oeuvre grammaticale, p. 144, n. 47.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 134

27-Sep-21 8:09:10 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

135

I. Formosa difformitas The hand that we may attribute to Alcuin or to his secretary was the one that sketched many of the manuscript’s drawings. That it is the same hand is ensured by two facts: the use, in some instances, of the same ink with which the corrections were made, and the presence of many abbreviations formed by letters written by the same person — “AL,” “ALC,” and the whole name “ALCUINUS” — within the same drawings, at times in the shape of a calligram.11 The drawings are often arranged along the lower margin, but sometimes also along the upper margin and along the left and right margins. At times, they are isolated images of men and animals, but the profile of one often overlaps with those of others, and the individual images almost seem to arise out of one another. But that is not all. Through a very peculiar technique aimed at disorienting the viewer by creating more than one optical illusion simultaneously, what often appears to be one image doubles or triples, since it contains within itself other smaller figures. This continuous trompe-l’oeil12 effect has an ancient origin, just as ancient as the method of incorporating and almost wedging bodies and faces into each other: in the frescoes, gems, candelabra, pottery, and common objects of Roman art, both classical and late antique, we find countless examples of this optical illusion.13 It is particularly developed in glyptics, above all in apotropaic, astrological, or magic-religious gems,14 in which it is often possible to find both figures merged with others, the so-called grylloi, and optical illusions based on pareidolia, such as faces that multiply themselves, but share one or more elements, or those that change before our eyes depending on whether we hold them straight or turn them upside-down (Plate 2 a, b). Similar optical illusions, although created through different techniques, lurk in calligrams.15 These are optical illusions in which the letters become 11 Troncarelli,

L’antica fiamma, pp. 136-152. Il trompe l’oeil. 13 Mulliez, Le luxe de l’imitation; J. Baltrušaitis, Il Medioevo fantastico: antichità ed esotismi nell’arte gotica; id., Risvegli e prodigi. La metamorfosi del gotico; id., Le miroir. On the calligrams and on the carmina figurata, see: Pozzi, La parola dipinta; Ernst, Carmen figuratum. 14 The literature on these topics is extensive. For a recent study, with an updated bibliography, see: Maioli, Magia e superstizione, pp. 99-111, in particular pp. 100-111 with bibliography; Sfameni, Tra religione e magia, pp. 377-340. On the evil eye and talismans against the evil eye, such as gems, see Elliott, Beware the Evil, IV. Finally, we note the essay by Coglitore, Pietre figurate. 15 Polara, Optaziano Porfirio, pp. 163-173. See also id., Aenigmata, pp. 197-216; id., Il Technopaegnion, pp. 277-284; id., Tra ars e ludus, pp. 55-70. 12 Calabrese,

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 135

27-Sep-21 8:09:10 PM

136

chapter v

images and the images letters, without anyone wondering what might be hidden “beneath the veil” of the appearance16 (Plate 2c). Alcuin, who preferred to be called Publius Alcuinus in honor of the renowned author of versus intexti, Publius Porfyrius Optatianus,17 cultivated the practice of the carmen figuratum with great diligence.18 Alcuin dedicated two of them to Charlemagne and directly inspired his disciple Rhabanus Maurus,19 with whom Alcuin is portrayed in the oldest and most authoritative manuscripts of his collection of visual poems, known as the Liber de laudibus Sanctae Crucis. Other authors gravitating around Charlemagne’s court, such as Theodulf, Paulinus of Aquileia, and Joseph Scotus, wrote poems of the same kind, composed of versus intexti.20 We encounter the same swarming of letters and bodies again in the drawings of the Codex Pagesianus, above all at the beginning and end of the volume. As if caught in the grip of an anguish caused by horror vacui, the artist covered the white spaces of the recto and the verso of the first leaf and the verso of the last one with a multitude of human faces, animal muzzles, and imaginary creatures. These drawings are later than Arbeo’s subscription, which is partially covered by some of the images, and prior to Leidrat’s subscription, which in turn covered some of the illustrations. The subjects portrayed, which are of a wide variety and clearly inspired by Roman art of the classical and late-antique period, mostly belong to the iconography of the grotesque images living in drôleries and in Romanesque and Gothic “grylloi” (hybrid figures generally showing a man with animal features or vice-versa)21 (Plates 36-52). These are the creatures “crowding” the margins of the Pagesianus: excited fauns, coarse faces, eager birds, men with ears like donkeys, harpies with hooked beaks, wolves, mice, basilisks ready to bite. This frantic crowd of shapeless figures is made even more dizzying by an obsessive game of optical illusions. A monster often contains within itself another monster, which reveals itself only for an instant 16 Spilling, Opus Magnentii; Rabani Mauri, In honorem sanctae crucis; Ferrari, Il “Liber sanctae crucis.” See also Guglielmetti, Hrabanus Maurus, in La trasmissione dei testi latini del medioevo. 17 Publii Optatiani Porfyrii Carmina, ed. G. P. Polara, I-II. On this topic, see: Deroux, The Carmina of Publilius Optatianus Porphyrius, pp. 447-466. 18 Chazelle, The Crucified God 19 Perrin, La poésie de cour carolingienne, pp. 333-335. 20 Ernst, Carmen figuratum, pp. 158-202. 21 Randall, Images; Nordenfalk, Drolleries, pp. 418-421; Demus, Pittura murale; Mel­ link ­ off, Riding Backwards, pp. 153-176; Schapiro, The Sculptures of Souillac, pp. 114-144); Sandler Reflections on the Construction of Hybrids, pp. 51-65; Wentersdorf, The Symbolic Significance, pp. 1-20; Camille, Image on the Edge; id., Mouths and Meanings, pp. 43-54; Mein ganz Körper; Wirth, Les marges à drôleries.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 136

27-Sep-21 8:09:10 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

137

Plate 36a-b – Roma, Museo Nazionale Romano, Gryllos, IInd-Ist century a.C. (Inv. 50/30 C; La glittica Santarelli ai Musei Capitolini. Intagli, cammei e sigilli, a cura di A. Gallottini, Roma 2012, n. 265); b) London British Museum, Gryllos, XVIIIth century (Inv. 1913,0307.100).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 137

27-Sep-21 8:09:11 PM

138

chapter v

Plate 36c – Raban, De laudibus S. Crucis, poem n. 3 (M. Perrin, L’iconographie de la Gloire à la sainte Croix de Raban Maur, Turnhout, 2000, pp. 128-129).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 138

27-Sep-21 8:09:12 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

139

Plate 37c – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 139

27-Sep-21 8:09:13 PM

140

chapter v

Plate 38a – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (details).

Plate 38b – Gryllos.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 140

27-Sep-21 8:09:13 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

141

Plate 39 – a) Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail); fe); b) Gryllos (Reading, Reading Museum, Duke of Wellington Collection, Inv.: 03006. cfr M. Henig, A corpus of Roman engraved gemstones from British sites, Oxford 1978 n. 377).

Plate 40 – a) British Museum, Drawing XVIIIth century, Inv. 2010,5006.1157 (BM Cat. Gem 2571); b) Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 141

27-Sep-21 8:09:15 PM

142

chapter v

Plate 41 – Details.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 142

27-Sep-21 8:09:16 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

143

Plate 42 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (details).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 143

27-Sep-21 8:09:17 PM

144

chapter v

Plate 43 – a) Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail); b) Marble relief, Blenheim Palace, Oxfordshire.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 144

27-Sep-21 8:09:18 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

145

Plate 44 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 145

27-Sep-21 8:09:19 PM

146

chapter v

Plate 45a – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail); b – Casale di Piazza Armerina (”Corridoio della grande caccia”), IVth century; c – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 146

27-Sep-21 8:09:19 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

147

Plate 46a-b – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail); c – Aquileia, Cripta (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 147

27-Sep-21 8:09:20 PM

148

chapter v

Plate 47a, c – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail); b, d – Palatino, Augustus Palace (1st c. B.C.).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 148

27-Sep-21 8:09:21 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

149

Plate 48a-b – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 149

27-Sep-21 8:09:21 PM

150

chapter v

Plate 49a – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail); b – Casa di Vila Graziosa, Musei Vaticani, sala dellle Nozze Aldobrandini (1st c. B.C.).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 150

27-Sep-21 8:09:23 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

151

Plate 50 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 151

27-Sep-21 8:09:23 PM

152

chapter v

Plate 51 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (details).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 152

27-Sep-21 8:09:24 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

153

Plate 52 – Biblioteca Apostolica Vaticana, Pagès 1, f. 1v (detail).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 153

27-Sep-21 8:09:25 PM

154

chapter v

after the first has been seen, such as, for example, the braying mule contained within the head of a roaring lion on f. 1v. At other times, we seem to catch sight of a single, disturbing shadow, but if one keeps his eyes open, voilà! a cluster of figures, such as the faun on f. 1v. The faun, instead of hair, has both the head of a wild boar and the head of a lion with its mane blowing in the wind; instead of a beard, another lion’s head; instead of cheeks, two devilish heads placed on top of one other; instead of a jawbone, two birds, one inserted into the other, the first with a pointed beak and the second with the face of a monster. The tension created by these effects is alleviated by the sudden appearance of some aerial entity, fictional and meek, such as the Zodiac Man, or by the appearance of some caricature smiling at us or provoking our smile, or by the appearance of some domestic animal suggesting a more quiet lifestyle, such as a lamb, cat, or dog. At times, we find ourselves isolating the letters with which an enigmatic face is interwoven, only to realize that those letters spell the same name over and over again. However, these letters also unexpectedly appear on the faces of demons and drag us back again into the Hell that we had previously departed. Soon, the indistinct parade of wild animals and monsters resumes and begins to weary our sight by parading foxes, stone martens, bears, sharks, snakes, and hungry dwarves before our eyes. Everything begins again, just like before, and we are made uncomfortable by it. The most interesting aspect of this complex sight is the technique used to draw the images. It is a technique that changes over and over again in a puzzling way. In some instances, the images are engraved onto the parchment with a sharp tool and proliferate before our eyes, almost escaping our attention, since we mistake them for the natural creases of wrinkled and rough parchment. In others, a very pale kind of ink is used that warns us about the presence of something poorly defined on the page. Nevertheless, what we perceive is unclear, and some indistinct figures are hard to recognize, both because of their color, which is barely visible, and because they consist of incomplete lines, of points detached from one another, and sometimes of slight smudges of ink that create shadows. In other instances, the opposite situation obtains: the images seem to have a position of prominence, because the surface of the parchment around them has been scraped with pumice. This allows the volume of those bodies to stand out, creating a contrast between the color of the parchment and that of the scraping. Moreover, in some cases, the figures have been drawn with white lead and stand out quite clearly against the dark background of the page. The result of this kaleidoscope of techniques is that the viewer is surprised by a kaleidoscope of images, not evident at first sight but rising to

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 154

27-Sep-21 8:09:25 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

155

the surface suddenly, emerging from a seemingly white background like a mysterious apparition. II. Imagines If we pass the first page of the Pagesianus and leaf through the entire manuscript, we can say that almost every page to which we direct our attention reveals figures similar to those described before, situated on the sides and lower margins. But, as noted before, we should not forget that these images are barely visible to the naked eye, and that only some of them (like those we saw before) are carefully drawn, whereas in most other instances they have been sketched very quickly, with the same type of haste used to jot down a marginal note following sudden inspiration. In truth, most of the images in the side margins make us think of something like an extemporary marginal annotation made while reading, playing the role of real, illustrated maniculae, similar to those small portraits used to point out meaningful passages. We find them in many medieval manuscripts, beginning with those occurring in the margins of Boccaccio’s autograph manuscripts.22 In the Pagesianus, the function of pointing out meaningful passages is performed by human profiles, with noses pointed toward a line of text, or by birds whose beaks are turned in the same direction. At times, this task is entrusted to the head of an animal, such as a horse, deer, dog, ox, or donkey, with its muzzle seeming to move in the direction of a passage that needs to be remembered (see, for example, ff. 8v, 26v, 44v). A somewhat similar function, namely, to help with the text, can be attributed to other images, both fictional and realistic, that crowded the white space at the end of a work. They seem to emphasize that there is a break between the previous text and the following one — sometimes for just a few lines and other times for a much longer distance — and that the reading has to be resumed beyond the space devoid of any alphabetical signs (e.g. ff. 30r-v). The drawings in the lower margins of the leaves are probably also meant to be a support for the reader, helping him to separate the writing area from the space below, considered a sacred area in which no one should be allowed to intervene. Finally, some less frequent imagines should be noted. They are reminiscent of the mere sketch of an illuminated initial that a professional illuminator will need to complete later. In these imagines, we find a precise echo of Late Antique art, both in each individual subject portrayed (Plate 22 Branca,

Boccaccio visualizzato; Cursi, La scrittura.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 155

27-Sep-21 8:09:25 PM

156

chapter v

Plate 53 – Biblioteca Apostolica Vaticana, Pagès 1, f. 113v (detail); b – Guernsey Museum at Candie, Guernsey (Roman collection, 2nd-3rd c. B.C.).

20), and, above all, in the superdekorativ taste, similar to that found in some sixth-century models which served as inspiration for Merovingian and Insular manuscripts. The images which recur, found in every possible position and serving all of the purposes already mentioned above, whether they be isolated, clinging to one another, or skillfully inscribed into one another, are the following: A human head, often with a grotesque expression the head of the Zodiac Man (with a Ram in his hair) the head of a faun the head of a man with ass-ears or with a hat which has ass-ears the heads of half-human and half-animal creatures the heads of animals or imaginary birds with hooked beaks or human faces the head of an ass or a mule the head of a horse the head of an elephant the head of a dog (of different breeds) the head of a wolf

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 156

27-Sep-21 8:09:25 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

157

the head of a rabbit or a whole rabbit (or a hare) the head of a he-goat or a whole he-goat the head of a sheep or a whole sheep the head of a roe deer or a whole roe deer the head of a deer the head of a fox the head of a hedgehog the head of a mouse or a whole mouse the head of a cat the head of a pig or wild boar the head of a bear the head of a large feline the head of a snake or a whole snake the head of a dragon the head of a large fish or a whole fish the head of an ox or bull the bucranium the head of a duck the head of a bird with a pointed beak We encounter the same images, or ones with similar shapes and symbolic meanings (often with the letters “A, L, C,” or “A, L, C, U, I, N, U, S” inserted into them), in some manuscripts associated in various ways with Alcuin. These manuscripts were written in Tours, in an environment in which the presence of Alcuin has been confirmed23 and dated, more or less by all scholars, to a period preceding Alcuin’s death: BAV, Reg. lat. 76224; Paris, BnF, lat. 884725; Troyes, Médiathèque du Grand Troyes, Fonds an­ cien 1742.26 It is necessary (and I intend to do it in the future) to study these 23 For all the manuscripts mentioned, see Rand, A Survey, pp. 96-97, 113-114; Bischoff, Katalog, III: Padua – Zwickau, Wiesbaden 2004, pp. 146-147, 386, 435; Ganz, Handschriften der Werke Alkuins, pp. 185-194. 24 Billanovich, Dal Livio di Raterio, pp. 26, 209; Vezin, La répartition du travail, pp. 213, 215; Reynolds,
 Livy, pp. 208-209; Bischoff, La biblioteca di corte di Carlo Magno, pp. 127-128; John, The Named, pp. 109, 111, 116, 117, 118, 119, 120; Von Büren, Livy’s Roman History, pp. 59, 60, 64, 88; Busonero, Alcune osservazioni, pp. 19-36. 25 Ferrari, Der älteste Touronische, pp. 108-113. 26 Coulson-Grotans, Classica Et Beneventana, p. 252; C. Cazaux-Kowlski, Une pièce négligée, p. 111-146, in particular pp. 133-134.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 157

27-Sep-21 8:09:25 PM

158

chapter v

manuscripts more carefully, and eventually other witnesses of the same kind, taking into account the chronological boundaries within which the study should be conducted.27 III. Latina Siren We have already noted en passant that behind this entire repertoire of shadows and illusions there stands Roman art, and, above all, Late Antique art with its grotesque drôleries and taste for the trompe l’oeil. We need only add that the influence of these models can be easily explained by attributing the authorship of the illustrations to Alcuin. As is well known, he was educated in York, a rich center of ancient manuscripts and culture, closely in touch with the neighboring monasteries of Lindisfarne and Wearmouth-Jarrow, in which extraordinary examples of Late Antique art were circulating, beginning with the Codex Grandior of Cassiodorus’ Bible. The future abbot of Tours would later go to Rome twice, spending a long period of time there. Alcuin thus had the chance to admire and quietly study the magnificent monuments of a past that had not yet disappeared, as well as prestigious manuscripts that no longer exist. In addition to this extraordinary experience, we cannot forget the importance of the Insular culture into which Alcuin had put down roots. It was a culture imbued with models inherited from Late Antiquity, although reinterpreted within an autonomous and innovative language. It may be enough to consider a masterpiece such as the Lindisfarne Gospels in order to understand how the unique echo of the Latin world was still clearly audible to those who wanted to hear it, even without a mythical journey to the distant Urbs. If the influence of certain sources is not hard to guess, it is somehow much more difficult to explain the nature of the secret inspiration that prompted a personality like Alcuin to pour such a large crowd of grotesque, bizarre, and fantastic figures onto the margins of a miscellany of philosophical texts that were considered important and rare in his environment. He doubtless wanted to pay homage to the art that he admired so much, and at the same time, to give free reign to his exuberant imagination, an imagination which could only surface rarely in his public life. And yet, 27 The

research should analyze the manuscripts dated before Alcuin’s death, keeping in mind, however, that the abbot of Tours began to have eye issues starting from 799 (Ep. 170) and that, consequently, it may have been hard for him to write and draw. It should be kept in mind, however, that the manuscripts dated by scholars after 799 may have been written before or, at any rate, may have been started long before the final date of their transcription (a case of which we also have instances in the manuscripts mentioned above: for example, in the Paris, BnF, lat. 8847 we find that the writing of the text covered preexisting drawings).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 158

27-Sep-21 8:09:25 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

159

faced with so many chimeras incessantly appearing, a suspicion arises. It seems that the devilish director of this theater of shadows has some secret meaning in mind, much more profound than playing an imaginative game for its own sake or signaling respect for a noble tradition. The same monsters, the same hybrids that were already popping out from between the lines of Insular manuscripts such as the Book of Kells and that proliferate in the Pagès manuscript will be reproduced in a short period of time and with much more energy on the columns and portals of the Romanesque churches, not only in scenes recalling Hell or the mob of damned souls, but also, and above all, in the spaces dedicated to the portrayal of the crowd of demons that lay siege to all Creation. This choice presupposes some view of the world and cannot be explained away as merely the expression of a wild imagination. After all, already in ancient and late-antique art, frightening masks, haunted expressions, misshapen creatures, human bodies generated from animal bodies, in frescoes, jewelry, objects, furniture, and glyptics, often hinted at something more intense and arcane than the fanciful game of a fine decoration made for its own sake. It is well known that in the ancient and late-antique world there was a deep-rooted belief that the universe was populated by a multitude of invisible beings, and that Neoplatonism and Orphism reinforced this primitive belief by evoking in various ways the presence of daimones scattered between heaven and earth, mediators between humanity and divinity. This opinion is effectively summarized by Apuleius that is worth quoting: “Moreover, there are certain divine middle powers, situated in this interval of the air, between the highest ether and earth, which is in the lowest place, through whom our desires and our deserts pass to the Gods. These are called by a Greek name daemons, who, being placed between the terrestrial and celestial inhabitants, transmit prayers from the one, and gifts from the other. They likewise carry supplications from the one, and auxiliaries from the other, as certain interpreters and saluters of both. Through these same daemons, as Plato says in the Banquet [202e-203a], all denunciations, the various miracles of enchanters, and all the species of presages, are directed. Prefects, from among the number of these, providentially attend to everything, according to the province assigned to each; either by the formation of dreams, or causing the fissures in entrails, or governing the flights of some birds, and instructing the songs of others, or by inspiring prophets, or hurling thunder, or producing the coruscations of lightning in the clouds; or causing other things to take place, by which we obtain a knowledge of future events. And it is requisite to think that all these particulars are effected by the will, the power, and authority of the celestial Gods, but by the compliance, operations, and ministrant offices

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 159

27-Sep-21 8:09:25 PM

160

chapter v

of daemons … and … all these, and other things of the like kind, properly accord with the middle nature of daemons. For as they are media between us and the Gods, in the place of their habitation, so likewise is the nature of their mind; having immortality in common with the Gods, and passion in common with the beings subordinate to themselves.”28 Against Apuleius came Augustine’s severe admonition, two-and-a-half centuries later. Augustine was the Christian theologian who was closest to Platonism, so much so that he considered it to be the best among the ancient philosophies and surely the closest to the truth. In the De Civitate Dei, the saint responded polemically to Apuleius, saying: “The same Apuleius, when speaking concerning the manners of demons, said that they are agitated with the same perturbations of mind as men; that they are provoked by injuries, propitiated by services and by gifts, rejoice in honors, are delighted with a variety of sacred rites, and are annoyed if any of them be neglected. Among other things, he also says that on them depend the divinations of augurs, soothsayers, and prophets, and the revelations of dreams, and that from them also are the miracles of the magicians. But, when giving a brief definition of them, he says, Demons are of an animal nature, passive in soul, rational in mind, aerial in body, eternal in time… Therefore it is no great thing for them to be of an animal nature, for so also are the beasts; in being rational as to mind, they are not above ourselves, for so are we also; and as to their being eternal as to time, what is the advantage of that if they are not blessed? For better is temporal happiness than eternal misery. Again, as to their being passive in soul, how are they in this respect above us, since we also are so, but would not have been so had we not been miserable? Also, as to their being aerial in body, how much value is to be set on that, since a soul of any kind whatsoever is to be set above everybody? And therefore religious worship, which ought to be rendered from the soul, is by no means due to that thing which is inferior to the soul.”29 Yet, despite Augustine’s censure, the old belief in the presence and power of the daemons remained undisturbed through the centuries that elapsed between the end of the ancient world and the modern era. This is proved by examples of every kind, from the huge number of talismans and gems possessing apotropaic value, in which Christian and pagan elements are easily combined, to the countless types of magic rituals full of Christian references and, above all, prayers to the angels to oppose the influence of demons, prayers which included both angels acknowledged by the Church 28 Apul., De deo Socr. 6-13 (trans. Th. Taylor). 29 De Civitate Dei VIII, 16 (trans. J. J. Smith).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 160

27-Sep-21 8:09:25 PM



The devil in the mirror: Alcuin’s drawings in the codex Pagesianus

161

and those that religious authorities declared extraneous to the Catholic faith. An obvious sign of inconsistency and ambiguity, even among the most thoughtful representatives of Medieval Christian culture, is found in the fact that these types of prayer, having the same value as talismans, were in some cases honored, recommended, and considered worthy of the highest respect by the very same personalities as abhorred the superstitious use of talismans among uneducated people. Alcuin was one of such personalities, ready to stigmatize the superstition of phylacteries among Jews and of necklaces with animal bones among uncouth English farmers. But Alcuin was also ready to praise and even use the apotropaic loricae widespread in the Celtic world, which in some cases were nothing but the same magic rituals — adapted to Christian doctrine — used to exorcize infirmities and the influence of malign spirits from the body.30 We should not be surprised by this manifestation of something that, in accordance with the history of medicine and cultural anthropology, we may call “spontaneous philosophy”31: “that heap of ideas and feelings, neither systematic nor formalized, in which elements of scientific knowledge and common sense, theoretical constructions and the results of experience, personal beliefs and the knowledge acquired through day-to-day activities converge.”32 The thousand-year-old figures of demons with apotropaic functions which spread along the boundaries between the inside and outside of houses and temples, like the margins of manuscripts, are rooted in the “spontaneous philosophy” of folkloric and learned traditions. This “spontaneous philosophy” is constantly strengthened by new philosophies which common sense adjusts and simplifies. The demons crowd in places out of which they can burst into our lives. Sometimes they are able to cross these thresholds, these boundaries, appearing before our eyes unexpectedly and unforeseen, as happens in mystery cults and, according to a classical and medieval tradition, during psychic experiences of “mediumship.” The latter can be detected through special objects which have exceptional powers, such as mirrors or any surface mirroring the sky and the atmosphere inhabited by mysterious spirits.33 30 Lambert, Celtic loricae, pp. 629-648. Patrick Sims-Williams believes, and rightly so, that Alcuin’s seeming inconsistency is a typical manifestation of a higher social class of High Medieval England that is opening up to post-Roman European civilization and does not want to have anything to do with the lifestyle and identity of a lower social class, still linked to a world of traditions (Sims-williams, Religion, pp. 273-327). 31 La città dei segreti, p. 18.
 32 Cosmacini, Scienza, p. 1224. 33 Baltrušaitis, Le miroir, pp. 188-203. See also The Mirror in Medieval and Early Modern

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 161

27-Sep-21 8:09:26 PM

162

chapter v

It is hard to establish the extent to which all of this is deliberately present in the drôleries of the Pagesianus manuscript, partially imitating ancient amulets, which seem to anticipate the grotesque figures of the Gothic period. Yet it is equally difficult to feign certainty about the opposite conclusion for the sole purpose of comforting ourselves, averting the influence of the fantastical figures, and thinking that they are just a game, the expression of the authors’ wild imagination. These elusive and vague images, carefully disguised in order to surprise the reader with their sudden appearance, are a sort of subliminal message that leaves a feeling of uneasiness. Perhaps, at least partially, along the margins of the manuscript that was supposed to attract the attention of a learned man and encourage him to exercise his reason, following the example of Aristotle, an ambiguous reflex, what Freud called the Unheim­ lich, was also hidden. It is the feeling that we are surrounded by something sinister and inexplicable, something that is perceived as both familiar and foreign to us, provoking anxiety and confusion. It is the devil’s grin in the mirror, sneering, disquieting, enigmatic.

Culture. There is a symmetrical correspondence between the millenary trust in apotropaic methods and the millenary trust in the existence of a network of evil and aggressive powers around us, which are directed against us, and whose quintessential representative is the evil eye.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 162

27-Sep-21 8:09:26 PM

BIBLIOGRAPHY Alcuin, Epistolae, edited by E. Dümmler, Berlin 1895. ———, Opera de novo collecta, emendata et aucta cura et studio F. Frobenii. Regensburg 1777. Ammannati, G. – Stagni, E., Ancora sul manoscritto Parigino latino 2769 (cc. 1-23) † 4808 (cc. 53-65), in Scrittura e Civiltà 25 (2001), pp. 399-424. Andreuccetti, P. A. – Lazzareschi Cervelli, I. (editors), Il colore nel Medioevo. Arte, simbolo, tecnica, Atti delle Giornate di studi, 2-4 Maggio 1996. Lucca 1998. Arbeo von Freising, in Lexikon des Mittelalters, I, München 1979. Asztalos, M., Boethius as a Transmitter of Greek Logic to the Latin West: the Categories, in Harvard Studies in Classical Philology 95 (1993), pp. 367-407. Baltrušaitis, J., Il Medioevo fantastico: antichità ed esotismi nell’arte gotica, Milan 1973. ———, Le miroir: essai sur une légende scientifique: révélations, science-fiction et fallacies, Paris 1978. ———, Lo specchio: rivelazioni, inganni e science-fiction, Milano 2007 (original edition: Le miroir: révélations, science-fiction et fallacies, Paris 1981). ———, Risvegli e prodigi. La metamorfosi del gotico, Milan 1999. Bartezzaghi, S., Incontri con la Sfinge. Nuove lezioni di enigmistica, Turin 2004. ———, Scrittori giocatori, Turin 2010. Beer, R., Monumenta palaeographica vindobonensia: Denkmäler der Schreibkunst aus der Handschriftensammlung des hapsburg-lotharingischen Erzhauses, Leipzig 1910. Billanovich, G., Dal Livio di Raterio (Laur. 63, 19) al Livio del Petrarca (B. M. Harl., 2493), in Itinera. Vicende di libri e testi, I, edited by M. Cortesi, Rome 2004, pp. 103-175. Bischoff, B., Die Hofbibliothek Karls des Grossen, in Karl der Große: Lebenswerk und Nachleben, II, Das geistige Leben, edited by B. Bischoff – W. Braunfels, Düsseldorf 1965. ———, Die südostdeutschen Schreibschulen und Bibliotheken in der Karolingerzeit, I, Die bayrischen Diözesen, Leipzig 1940. ———, Frühkarolingische Handschriften und ihre Heimat, in Scriptorium 23 (1968), pp. 306-314. ———, Karl der Grosse: Werk und Wirkung [Ausstellung], Aachen 1965. ———, Katalog der festländischen Handschriften des neunten Jahrhunderts (mit Aus­ nahme der wisigotischen), II, Laon – Paderborn, Wiesbaden 2004. ———, La biblioteca di corte di Carlo Magno, in Le biblioteche nel mondo antico e medievale, edited by G. Cavallo, Bari – Rome 1988, pp. 113-135.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 163

27-Sep-21 8:09:26 PM

164

bibliography

———, Mittelalterliche Studien. Ausgewählte Aufsätze zur Schriftkunde und Litera­ turgeschichte. Stuttgart 1981. ———, Paleografia latina. Antichità e Medioevo, edited by G. P. Mantovani – S. Zamponi, Padova 1992. ———, Panorama der Handschriftenüberlieferung aus der Zeit Karls des Großen, in Karl der Grosse. Lebenswerk und Nachleben, 3, edited by Helmut Beumann – Wolfgang Braunfels, Düsseldorf 1965-1968, pp. 233-254. ———, Scheda n° 417, in Codices Latini Antiquiores, 4, edited by E. A. Lowe, Oxford 1947. Bitterauf, T., Die Traditionen des Hochstifts Freising. München 1905. Boethius, Commentarii in librum Aristotelis ΠΕΡΙ ΕΡΜΗΝΕΙΑΣ, edited by C. Meiser, Leipzig 1877-1880. Boulhol, p. Claude de Turin. Un êvèque iconoclaste dans l’Occidente carolingien. Paris 2002. Braakhuis, H. A. G. – Kneepkens, C. H. (editors), Aristotle’s Peri hermeneias in the Latin Middle Ages: Essays on the Commentary Tradition. Turnhout 2003. Branca, V. (editor), Boccaccio visualizzato. Narrare per parole e per immagini tra Medioevo e Rinascimento. Turin 1999. Bullough, D., Alcuin and the Kingdom of Heaven: Liturgy, Theology and the Caro­ lingian Age, in id., Carolingian Renewal: Sources and Heritage, Manchester 1991. ———, Charlemagne’s Court Library Revisited, in Early Medieval Europe 12 (2003), pp. 339-363. Burn, A. E. (editor), Facsimiles of the Creeds from Early Manuscripts. Edited by A. E. Burn, D.D. with palaeographical notes by the late Dr. Ludwig Traube. London 1909. Busonero, P., Alcune osservazioni sul rapporto tra antigrafo e apografo nella tradi­ zione della terza decade di Livio, in Nuovi annali della Scuola speciale per archi­ visti e bibliotecari 18 (2004), pp. 19-36. Caffaro, A., Scrivere in oro. Ricettari medievali d’arte e d’artigianato (secoli IX-XI). Codici di Lucca e di Ivrea. Naples 2003. Calabrese, O., Il trompe l’oeil. Milan 2013. Cameron, A. D. E., The Last Pagans of Rome. Oxford 2011. Camille, M., Image on the Edge. The Margins of Medieval Art. London-Cambridge, Mass. 1992 id., Mouths and Meanings: Towards an Anti-Iconography of Medieval Art, in Iconography at the Crossroads, edited by B. Cassidy, Princeton 1993, pp. 43-54. Cappelli, A., Dizionario delle abbreviature latine ed italiane, edited by M. Geymonat – F. Troncarelli, Milano 2011. Cazaux-Kowlski,Cazaux-Kowlski, C., Une pièce négligée du puzzle neumatique français: la notation lyonnaise, in Solange Corbin et les débuts de la musicologie médiévale, edited by C. Cazaux-Kowlski – I. His – J. Gribenski, Rennes 2015, pp. 111-146. Cennini, C., Libro dell’Arte. Vicenza 1975.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 164

27-Sep-21 8:09:26 PM

bibliography

165

Chadwick, H., Boethius: The Consolations of Music, Logic, Theology, and Philoso­ phy, Oxford 1981. Champion, M., Ill-Wishing on the Walls: Late Medieval Graffiti Curses from Norwich Cathedral, in Norfolk Archaeology XLVI (2014), pp. 61-66. ———, Medieval Graffiti Inscriptions Found in All Saints Church, Litcham, in Nor­ folk Archaeology XLVI (2011), pp. 199-208. Chazelle, F., The Crucified God in the Carolingian Era: Theology and Art of Christ’s Passion, Cambridge 2001. Coglitore, R., Pietre figurate Forme del fantastico e mondo minerale, Pisa 2004. Concilia Toletana III, in Collectio canonum ecclesiae Hispanae, Madrid 1808. Cornelii a Lapide, Commentarius in Pentateuchum Moysis, Venice 1627. Cosmacini, G., Scienza e ideologia nella medicina del Novecento: dalla scienza egemo­ ne alla scienza ancillare, in Annali della storia d’Italia, 7, edited by Della Peruta, F., Turin 1984, pp. 1223-1268. Costamagna, G. – Baroni, F. – Zagni, L., Notae tironianae quae in lexicis et in char­ tis reperiuntur novo discrimine ordinatae, Rome 1983. Coulson, F. T. – Grotans, A. A. (editors), Classica et Beneventana: Essays Presented to Virginia Brown on the Occasion of Her 65th Birthday, Turnhout 2008. Courcelle, P., Histoire d’un brouillon Cassiodorien, in Revue des études anciennes 44 (1942), pp. 65-86; reprinted in id., Opuscula selecta, Paris 1984, pp. 77-98. ———, Late Latin Writers and Their Greek Sources, translated by H. E. Wedeck, Cambridge, Mass. 1969. Crivello, F. – Denoël, F. – Orth, P. (editors), Das Godescalc Evangelistar. Eine Prachthandschrift für Karl den Großen, Darmstadt 2011. Cuppo Csaki, L., Pontifices di Costantinopoli nel Liber Pontificalis del settimo secolo: note sul codice Biblioteca Apostolica Vaticana, Vat. lat. 3764, in Rivista di storia e letteratura religiosa 43 (2007), pp. 359-372. Cursi, M., La scrittura e i libri di Giovanni Boccaccio, Rome 2013. Dales, D., Alcuin: His Life and Legacy, Cambridge 2012. ———, Alcuin: Theology and Thought, Cambridge 2013. De Gregorio, G., Tardo Medioevo greco latino: manoscritti bilingui d’Oriente e d’Oc­ cidente, in Libri, documenti, epigrafi medievali: possibilità di studi comparativi, Atti del convegno internazionale di studio dell’Associazione Italiana dei Paleografi e Diplomatisti, (Bari 2-5 ottobre 2000), edited by F. Magistrale – P. Fioretti – C. Drago, Spoleto 2002, pp. 43-44. Dekkers, E. (editor), Clavis Patrum latinorum, Turnhout 1995-2010. Delisle, L., Mèmoire de l’Ecole calligraphique de Tours, Paris 1885. ———, Notices sur plusieurs anciens manuscrits de la Bibliothèque de Lyon, in No­ tices et extraits 29 (1880), pp. 390-391. ———, Notice sur un manuscrit de l’église de Lyon du temps de Charlemagne, in No­ tices et extraits des manuscrits de la Bibliothéque Nationale et autres bibliothèques 35 (1897), pp. 831-842. Demus, O. Pittura murale romanica, Milan 1969.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 165

27-Sep-21 8:09:26 PM

166

bibliography

Denoël, C., Entre imitation et invention: un livre d’Évangiles de style tourangeau, in Les manuscrits carolingiens, edited by J.-P. Caillet – M.-P. Laffitte, Turnhout 2009, pp. 89-120. ———, La vallée de la Loire: de Saint-Martin de Tours à Fleury, in Trésors carolin­ giens. Livres manuscrits de Charlemagne à Charles le Chauve, edited by M.-P. Laffitte – F. Denoël, Paris 2007, pp. 151-165. Desnoyés, J. – Delisle, L., Note sur un monogramme d’un prêtre artiste du IXe siècle, in Comptes rendues des séances de l’Academie des Inscriptions et BellesLettres 30 (1886), pp. 376-381. D’Onofrio, G. (editor), Excerpta Isagogarum et Categoriarum, Turnhout 1996. ———, Fons scientiae: la dialettica nell’Occidente tardo-antico, Naples 1986. id. (editor), Vera Philosophia: Studies in Late Antique, Early Medieval and Renais­ sance Christian Thought, Turnhout 2008. Dopsch, H. – Juffinger, R., Virgil von Salzburg: Missionar und Gelehrter, Salzburg 1984. Dümmler, E. (editor), Poetae latini aevi Carolini, Berlin 1881. Edwards, J. S., The Carmina of Publilius Optatianus Porphyrius and the Creative Process, in Studies in Latin Literature and Roman History, edited by C. Deroux, Brussels 2005, pp. 447-466. Einhard, Vita Karoli, edited by P. Chiesa, Firenze 2014. Elliott, J. H., Beware the Evil Eye: The Evil Eye in the Bible and the Ancient World. Eugene, Oregon – Cambridge 2016. Ernst, U., Carmen figuratum. Geschichte des Figurengedichts von den antiken Ur­ sprüngen bis zum Ausgang des Mittelalters. Köln 1991. Faraone, A. – Kopp, A., Inversion, Adversion and Perversion as Strategies in Latin Curse Tablets, in Magical Practice in the Latin West: Papers from the International Conference Held at the University of Zaragoza, 30 Sept. – 1 Oct. 2005, edited by R. L. Gordon – F. M. Simón, Leiden – Boston 2010, pp. 381-398. Ferrari, M. C., Der älteste Touronische Pandekt Paris, Bibliothèque Nationale de France 8847 und seine Fragmente, in Scriptorium 53 (1999), pp. 108-113. ———, Il “Liber sanctae crucis” di Rabano Mauro. Testo, immagine, contesto. Bern 1999. Fisquet, H., La France Pontificale (Gallia Christiana). Histoire chronologique et bio­ graphique des archevêques et evêques des tous les diocèses de France. Paris 1864. Freeman, A. – Meyvaert, P. (editors), Opus Caroli Regis contra synodum (Libri Carolini), Hannover 2003. Frelick, N., (editor) The Mirror in Medieval and Early Modern Culture: Specular Reflections, Turnhout 2016. Fried, J., Karl der Große. Gewalt und Glaube. Eine Biographie, München 2014. Galonnier, A., Anecdoton Holderi ou Ordo generis Cassiodororum: Éléments pour une Étude de l’authenticité boécienne des Opuscula sacra, Louvain – Paris 1997 (Cited as Anecdoton).

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 166

27-Sep-21 8:09:26 PM

bibliography

167

———, Anecdoton Holderi ou Ordo generis Cassiodororum: Introduction, édition, traductionet commentaire, in Antiquité tardive 4 (1997), pp. 299-312 (Cited as Ordo generis). Ganz, D., Corbie in the Carolingian Renaissance. Sigmaringen 1990. ———, Handschriften der Werke Alkuins aus dem 9. Jahrhundert, in Alkuin von York und die geistige Grundlegung Europas. Akten der Tagung vom 30. September bis zum 2. Oktober 2004 in der Stiftsbibliothek St. Gallen, edited by E. Tremp – K. Schmuki, Saint Gall 2010, pp. 185-194. Garand, M.-F. – Mabille, M. – Muzerelle, D., Catalogue des manuscrits en écri­ ture latine portant des indications de date, de lieu ou de copiste. IV/1: Biblio­ thèque nationale, fonds latin (supplément), Nouvelles acquisitions latines, petits fonds divers. Notices établies par M.-F. Garand, M. Mabille et D. Muzerelle, avec le concours de M.-T. d’Alverny, Paris 1981. Gardthausen, V., Das alte Monogram, Leipzig 1924. Gemeinhardt, P., Die Filioque-Kontroverse zwischen Ost- und Westkirche in Früh­ mittelalter, Berlin 2002. Giuliani, E. – Pavolini, C., La ‘Biblioteca di Agapito’ e la Basilica di Sant’Agnese, in The Transformations of Urbs Roma in Late Antiquity, edited by W. V. Harris, Portsmouth 1999, pp. 85-108. Glaser, H. – Brunhölzl, F. – Benker, S. (editors), Vita Corbiniani: Bischof Arbeo von Freising und die Lebensgeschichte des hl. Korbinian, München 1983. Glauche, G., Bistum Freising, in Mittelalterliche Bibliothekskatalogen Deutschland und der Schweiz, IV, 2, München 1979. ———, Katalog der lateinischen Handschriften der Bayerischen Staatsbibliothek München. Die Pergamenthandschriften aus dem Domkapitel Freising, 1. Wiesbaden 2000. Gorman, M. M., The Myth of Hiberno-Latin Exegesis, in Revue bénédictine 110 (2000), pp. 42-85. Green, R. P. H., Latin Epics of the New Testament: Juvencus, Sedulius, Arator. Oxford 2006. Gresham, C. A., Medieval Stone Carving in North Wales: Sepulchral Slabs and Effi­ gies, Cardiff 1968. Gruber, J., Boethius: Eine Einführung, Stuttgart 2011. Guarducci, M., Scritti scelti sulla religione greca e romana e sul cristianesimo, Leiden 1983. Guglielmetti, R. E., Hrabanus Maurus, in La trasmissione dei testi latini del me­ dioevo. Mediaeval Latin Texts and Their Transmission, 3, edited by P. Chiesa – L. Castaldi, Florence 2008, pp. 275-332. Hadot, P., Marius Victorinus: Recherches sur sa vie et ses oeuvres, Paris 1971. Hofstadter, D. R., Ambigrammi. Un microcosmo ideale per lo studio della creati­ vità. Firenze 1987. Holtz, L., La minuscule marginale et interlinéaire de Florus de Lyon, in Gli Auto­ grafi medievali: Problemi paleografici e filologici. Atti del convegno di studio della

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 167

27-Sep-21 8:09:26 PM

168

bibliography

Fondazione Ezio Franceschini, Erice, 25 settembre – 2 ottobre 1990, edited by P. Chiesa – L. Pinelli, Spoleto 1994. ———, Leidrat, évêque de Lyon (798-814): ses livres, son écriture, in Amicorum so­ cietas: mélanges offerts à François Dolbeau pour son 65e anniversaire, edited by J. Elfassi – C. Lanéry – A.-M. Turcan-Verkerk, Firenze 2013, pp. 315-334. ———, L’évêque, le diacre et le manuscrit. Un grand moment de l’histoire du livre à Lyon: la bibliothèque épiscopale du IXe siècle, in Gryphe. Revue de la Bibliothèque de Lyon 5 (2002), pp. 28-34. ———, L’oeuvre grammaticale d’Alcuin dans le contexte de son temps, in Alkuin von York und die geistige Grundlegung Europas. Akten der Tagung vom 30. September bis zum 2. Oktober 2004 in der Stiftsbibliothek St. Gallen, edited by E. Tremp – K. Schmuki, Saint Gall 2010, pp. 129-149. Hunt, T., Popular Medicine in Thirteenth-century England: Introduction and Texts, Cambridge 1990. John, J. J., The Named (and Namable) Scribes in Codices Latini Antiquiores, in Scribi e colofoni. Le sottoscrizioni di copisti dalle origini all’avvento della stampa, Atti del seminario di Erice, X Colloquio del Comité international de paléographie latine (23-28 ott. 1993), edited by E. Condello – G. De Gregorio, Spoleto 1995, pp. 107-121. Jullien, M.-H., Alcuin et l’Italie, in Annales de Bretagne et des Pays de l’Ouest 111113 (2004), pp. 393-406. Jullien, M. H. – Perelman, F. (editors), Clavis des auteurs latins du Moyen Age. Territoire français 735-987, Turnhout 1994. Keefe, S. A., A Catalogue of Works Pertaining to the Explanation of the Creed in Carolingian Manuscripts, Turnhout 2012. Klibansky, R. – Regen, F., Die Handschriften der philosophischen Werke des Apu­ leius. Ein Beitrag zur Überlieferungsgeschichte, Göttingen 1993. Krämer, S., Bibliographie Bernhard Bischoff und Verzeichnis aller von ihm herange­ zogenen Handschriften, Frankfurt am Main 1998. Kristeller, P. O., Iter Italicum, 6, London – Leiden – Kobenavhn – Köln 1992. Kroll, K. – Steger, H. (editors), Mein ganzer Körper ist Gesicht: groteske Dar­ stellungen in der europäischen Kunst und Literatur des Mittelalters, Freiburg im Breisgau 1993. Laffitte, M.-P. – Denoël, C. – Besseyre, M. (editors), Trésors carolingiens. Livres manuscrits de Charlemagne à Charles le Chauve, Paris 2007. Lambert, P.-Y., Celtic loricae and ancient magical charms, in Magical Practice in the Latin West. Papers from the International Conference Held in Zaragoza, 30 Sept.-1 October 2005, edited by R. L. Gordon – F. M. Simón, Leiden – Boston 2010, pp. 629-648. Lapidge, M., Autographs of Insular Latin Authors of the Early Middle Ages, in Gli autografi medievali Gli autografi medievali. Problemi paleografici e filologici. Atti del convegno di studio (Erice, 25 settembre – 2 ottobre 1990), edited by P. Chiesa – L. Pinelli, Spoleto 1994, pp. 103-136.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 168

27-Sep-21 8:09:27 PM

bibliography

169

Lapidge, M. – Garfagnini, G. C. – Leonardi, C. – Santi, F. et Al. (editors), Com­ pendium auctorum Latinorum Medii Aevi (500-1500), Firenze 2000. Lauer, P., Observations sur le Scriptorium de Lyon, in Bibliothèque de l’Ecole des Chartes 86 (1925), pp. 380-387. Lebecq, S., Marchands et navigateurs frisons du haut Moyen Âge: Corpus des Sources écrites, Lille 1983. Leist, F. Urkundenlehre: Katechismus der Diplomatik, Paläographie, Chronologie und Sphragistik, Leipzig 1893. Lot, F. – Lauer, P. (editors), Diplomata Karolinorum, Paris – Toulouse 1941. Lowe, E. A. Codices Lugdunenses antiquissimi. Le scriptorium de Lyon, la plus an­ cienne école calligraphique de France, in Documents paléographiques, typographiques, iconographiques de la Bibliothèque de Lyon, fasc. 3-4, Lyon 1924. ———, Nugae Palaeographicae, in Palaeographical Papers 1907-1965, edited by L. Bieler, Oxford 1972, pp. 315-325. Maioli, M. G., Magia e superstizione, in Immagini divine. Devozioni e divinità nella vita quotidiana dei Romani edited by J. Ortalli – D. Neri, Florence 2007, pp. 99-112. Maag, N., Alemannische Minuskel (744-846 n. Chr.). Frühe Schriftkultur im Boden­ seeraum und Voralpenland, Stuttgart 2014. Macalister, R. A. S., Corpus Inscriptionum Insularum Celticarum, II, Dublin 1949. ———, The “Druuides” Inscription at Killeen Cormac, County Kildare, in Proceedings of the Royal Irish Academy, Section C: Archaeology, Celtic Studies, History, Lin­ guistics, Literature 32 (1914-1916), pp. 227-238. ———, The History and Antiquities of Inis Cealtra, in Proceedings of the Royal Irish Academy 33 (1916), pp. 93-174. Machielsen, J., Clavis patristica pseudepigraphorum Medii Aevi, Turnhout 1990. Magee, J., (editor) Anicii Manlii Severini Boethii De divisione liber, Leiden 1998. ———, and J. Marenbon, Appendix, in The Cambridge Companion to Boethius, edited by J. Marenbon, Cambridge 2009, pp. 303-310. ———, Boethius, in The Cambridge History of Philosophy in Late Antiquity, II, edited by L. Gerson, Cambridge 2010, pp. 788-812. ———, Boethius’ Consolatio and Plato’s Gorgias, in Boethius as a Paradigm of Late Ancient Thought, edited by T. Böhm – T. Jürgasch – A. Kirchner, Berlin – Boston 2014, pp. 13-29. ———, On the Composition and Sources of Boethius’ Second Peri Hermeneias Com­ mentary, in Vivarium 48 (2010), pp. 7-54. ———, Preliminary Observations on the Textual Tradition of Boethius’ First Peri Hermeneias Commentary, in Logic and Language in the Middle Ages: A Vol­ ume in Honour of Sten Ebbesen, edited by J. L. Fink – H. Hansen – A. M. Mora-Márquez, Leiden 2013, pp. 13-25. ———, The Text of Boethius’ De divisione, in Vivarium 32 (1994), pp. 1-50.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 169

27-Sep-21 8:09:27 PM

170

bibliography

Magrini, S., Un poco noto frammento altomedievale del De divisione di Boezio (Ver­ celli, Biblioteca Capitolare, CXXXI), in Per Gabriella: Studi in recordo di Gabriella Braga, III, edited by M. Palma – C. Vismara, Cassino 2013, pp. 1151-67. Marenbon, J., From the Circle of Alcuin to the School of Auxerre: Logic, Theology and Philosophy in the Early Middle Ages, Cambridge 1981. ——— (editor), The Cambridge Companion to Boethius, Cambridge 2009. ———, Tradition of Logic. Early Medieval Philosophy (480-1150): An Introduction, London – New York 2002. Martianus Capella, De Nuptiis Philologiae et Mercurii, edited by F. Eyssenhardt, Leipzig 1866. Martin, J. B., Invéntaire méthodique des manuscrits conservés dans les bibliothèques privées de la région Lyonnaise, in Revue des bibliothèques 7 (1897), pp. 486-487. ———, Note sur quelques ouvrages lyonnais rares ou inconnus, in Bulletin historique et philologique du comité des travaux historiques et scientifiques 47 (1897), pp. 119-120. McKitterick, R., Charlemagne: The Formation of a European Identity, Cambridge 2008. McManus, D., A Guide to Ogam, Maynooth 1991. Mellinkoff, R., Riding Backwards: Theme of Humiliation and Symbol of Evil, in Viator 4 (1973), pp. 153-176. Meyvaert, P., Medieval notions of publication: The “unpublished” “Opus Caroli regis contra Sinodum” and the Council of Frankfort (794), in Journal of Medieval Latin 12 (2002), pp. 78-89. Minio-Paluello, L., (editor) Aristoteles latinus, I, II, V, Bruges – Paris 1961, 1965, 1969. ———, Aristoteles latinus. Supplementa altera, Paris 1961. ———, Note sull’Aristotele latino medievale, X, I “Topici” nel X-XI secolo: Due fogli del testo perduto della redazione Boethiana definitiva. – Altri frammenti dai libri I-VI e VIII della medesima redazione, in Rivista di Filosofia Neo-Scolastica 50 (1958), pp. 97-111; reprinted in id., Opuscula: The Latin Aristotle, Amsterdam 1972, pp. 357-376. ———, Nuovi impulsi allo studio della logica: la seconda fase della riscoperta di Ari­ stotele e di Boezio, in La scuola nell’Occidente latino dell’Alto Medioevo, II, Spoleto 1972, pp. 743-766. ———, Opuscula: the Latin Aristotle, Amsterdam 1972. Mordek, H., Leidrat, Bischof von Lyon (797-814/6), in Lexikon des Mittelalters, 5, München – Zürich 1991, p. 1855. Mulliez, M., Le luxe de l’imitation: les trompe-l’oeil de la fin de la République ro­ maine, mémoire des artisans de la couleur, Naples 2014. Munk Olsen, B., Chronique des manuscrits classiques latins (IXe-XIIe siècle), III, in Revue d’histoire des textes 27 (1997), pp. 29-85. ———, L’étude des auteurs clasiques latins aux 11. et 12. siècles, 4.2, La réception de la littérature classique: manuscrits et textes, Paris 2014.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 170

27-Sep-21 8:09:27 PM

bibliography

171

Nordenfalk, C., Drolleries, in The Burlington Magazine 109 (1967), pp. 418-421. Pecere, O., Cassiodoro e la protostoria di un ‘corpus’ di scritti di Boezio, in Segno e Testo 12 (2014), pp. 149-221. ———, Modalità compositive e circolazione privata del libro nel tardoantico: il caso di Boezio, in Segno e Testo 13 (2015), pp. 219-31. Perrin, M. J., La poésie de cour carolingienne, les contacts entre Alcuin et Hraban Maur et les indices de l’influence d’Alcuin sur l’In honorem sanctae crucis, in Annales de Bretagne et des Pays de l’Ouest 111 (2004), pp. 333-351. Petrie, G., Christian Inscriptions in the Irish Language. Cambridge 2015. Petrucci, A., Breve storia della scrittura latina. Nuova edizione riveduta e aggiornata, Rome 1989. ———, Le biblioteche antiche, Turin 1983. Polara, G., Aenigmata, in Lo spazio letterario del Medioevo, 1. Il Medioevo latino, I.2, Rome 1993, pp. 197-216. ———, Il Technopaegnion di Ausonio: «Ludus laboranti», in Giornale italiano di filologia 49 (1997), pp. 277-284. ———, Optaziano Porfirio tra il calligramma antico e il carme figurato di età medio­ evale, in Invigilata lucernis 9 (1987), pp. 163-173. ———, Tra ars e ludus. Tecnica e poetica del Technopaegnion di Ausonio, in id., (editor) La poesia tardoantica e medievale, Alessandria 2001, pp. 55-70. Pozzi, G., La parola dipinta, Milano 1981. ———, Poesia per gioco. Prontuario di figure artificiose, Bologna 1984. Pratesi, A. – Cherubini, P., Paleografia latina: l’avventura grafica del mondo occi­ dentale, Vatican City 2010. Publilius Optatianus Porfyrius, Publilii Optatiani Porfyrii Carmina, edited by G. P. Polara, Turin 1973. Rabanus Maurus, In honorem sanctae crucis, edited by M. Perrin, Turnhout 1997. Radiciotti, P., Romania e Germania a confronto: un codice di Leidrat e le origini medievali della minuscola carolina, in Scripta 1 (2008), pp. 121-144. Rand, E. K., A Survey of the Manuscripts of Tours, I, Cambridge, Mass. 1929. Randall, L. M. C., Images in the Margins of Gothic Manuscripts, Berkley – Los Angeles 1966. Raniero Giordani da Pisa († 1348), Pantheologia emendata et aucta, I, Lyons 1655. Reynolds, L. D., Livy, in Texts and Transmission: A Survey of the Latin Classics, edited by id., Oxford 1984, pp. 208-209. Sandler, L. F., Reflections on the Construction of Hybrids in Gothic Marginal Illus­ tration, in Art, the Ape of Nature. Studies in Honor of W. H. Janson, edited by M. Barasch – L. F. Sandler, New York 1981, pp. 51-65. Sawyer, P., The Wealth of Anglo-Saxon England, Oxford 2013. Scalon, C. – Pani, L., Le Alpi porta d’Europa: scritture, uomini, idee da Giustiniano al Barbarossa. Atti del convegno internazionale di studio dell’Associazione Italiana dei Paleografi e Diplomatisti, Cividale del Friuli (5-7 ottobre 2006), Spoleto 2009.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 171

27-Sep-21 8:09:27 PM

172

bibliography

Schapiro, M., The Sculptures of Souillac, in id., Romanesque Art, London 1977, pp. 102-130 (Italian translation Arte romanica, Turin 1982, pp. 114-144). Schmitz, W., Commentarii notarum tironianum cum prolegomenis adnotationibus criticis et exegeticis notarumque indice alphabetico, Leipzig 1893. Sfameni, C., Tra religione e magia: temi isiaci nelle gemme di età imperiale, in Isis en Occident Actes du IIème Colloque international sur les études isiaques, Lyon III 16-17 mai 2002, Leiden 2003, pp. 377-340. Siegmund, I., Die Überlieferung der griechischen christlichen Literatur in der latei­ni­ schen Kirche bis zum 12. Jahrhundert, München 1949. Sims-Williams, P., Religion and Literature in Western England, 600-800, Cambridge 1990. Spilling, H., Opus Magnentii Hrabani Mauri in honorem sanctae crucis conditum. Hrabans Beziehung zu seinem Werk, Frankfurt 1992. Stegmüller, F., Repertorium Biblicum Medii Aevi, Madrid 1950-1980. Stengel, G. – Reihing, J., Nova S. Scripturae Tinea, Ingolstadt 1624. Stoppacci, P., Una silloge per tre maestri: il codex Renati e la dialectica Boethii tra Gerberto d’Aurillac, Abbone di Fleury e Fulberto di Chartres, in Mittellateinisches Jahrbuch 52 (2017), pp. 1-30. Suckale-Redlefsen, G., Die Handschriften des 8. bis 11. der Staatsbibliothek Bam­ berg, Wiesbaden 2004. Supino Martini, P., Aspetti della cultura grafica a Roma fra Gregorio Magno e Grego­ rio VII, in Scritti romani: scrittura, libri e cultura a Roma in età medievale, edited by G. Ancidei – E. Condello – M. E. Malavolta – L. Miglio – M. Signorini – C. Tedeschi, Roma 2012, pp. 213-250. Tafel, S., The Lyons Scriptorium, in Palaeographia latina 2 (1923), pp. 66-73 and 4 (1925), pp. 40-70. Tischler, M., Einharts Vita Karoli: Studien zur Entstehung, Überlieferung und Rezep­tion, Hannover 2001. Trithemius, J. Liber de scriptoribus ecclesiasticis, Sponheim 1492. Troncarelli, F. P., Afterword, in A Companion to Boethius in the Middle Ages, edited by N. H. Kaylor – P. E. Phillips, Leiden 2012, pp. 519-49. ———, Boezio a Costantinopoli: testi, contesti, edizioni, in Litterae Caelestes 3 (20089), pp. 191-225. ———, Cogitatio mentis: l’eredità di Boezio nell’Alto Medioevo, Naples 2005. ——— (editor), La città dei segreti. Magia, astrologia e cultura esoterica a Roma (XVXVIII secolo), Milan 1985. ———, L’antica fiamma: Boezio e la memoria del sapere antico nell’alto medioevo, Rome 2017. ———, Tradizioni perdute: la Consolatio philosophiae nell’alto medioevo, Padua 1981. ———, Umanesimo tardoantico: l’ultimo dei Romani e la consolazione della saggez­ za, Rome 2012.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 172

27-Sep-21 8:09:27 PM

bibliography

173

———, Una pietà più profonda. Scienza e medicina nella cultura monastica medie­ vale italiana, in Dall’eremo al cenobio, edited by G. Pugliese Carratelli, Milan 1987, pp. 703-727. ———, Vivarium: I libri, il destino. Turnhout 1998. Van De Vyver, A., Les étapes du développement philosophique du Haut Moyen-Âge, in Revue belge de philologie et d’histoire 8 (1929), pp. 425-452. Vezin, J., La réalization materielle des manuscrits latins pendant le Moyen Âge, in Codicologica 2 (1978), pp. 15-51. ———, La répartition du travail dans les «scriptoria» carolingiens, in Journal des Savants 3 (1973), pp. 212-227. ———, Manuscrits présentant des traces de l’activité en Gaule de Théodulphe d’Or­ léans, Claude de Turin, Agobard de Lyon et Prudence de Troyes, in Coloquio sobra circulación de códices y escritos entre Europa y la Península en los siglos VIIIXIII entre Europa y la Península en los siglos VIII-XIII. Actas, 16-19 sept. 1982, Santiago de Compostela 1988, pp. 157-171. Vitiello, M. Accusarentur saecula, si talis potuisset latere familia: Il fantasma de Severino Boezio nell’Italia dei Goti, in Historia 60 (2011), pp. 343-82. ———, Amalsuintha: The Transformation of Queenship in the Post-Roman World, Philadelphia 2017. ———, Cassiodorus anti-Boethius?, in Klio 90 (2008), pp. 461-84. ———, Sulle tracce di Patrizio, un amico di Boezio, in Hermes 138 (2010), pp. 23049. ———, Theodahad: A Platonic King at the Collapse of Ostrogothic Italy, Toronto – Buffalo – London 2014. Von Büren, V., Livy’s Roman History in the Eleventh-Century Catalogue from Cluny: The Transmission of the First and Third Decades, in Medieval Manuscripts of the Latin Classics: Production and Use, Proceedings of the Seminar in the History of the Book to 1500, edited by C. A. Chavannes Mazel – M. McFadden Smith-­ Anderson, Leiden 1993, pp. 57-73. Ward, A., Codex Pagesianus, in Codices Boethiani: A Conspectus of Manuscripts of the Works of Boethius, 3, Italy and the Vatican City, 3, edited by L. Smith – M. Passalacqua – S. Magrini – V. Longo, London – Turin 2001, pp. 309-311. ———, The Codex Pagesianus: Witness to Church Renewal, in The American Bene­ dictine Review 44 (1993), pp. 303-333. Wentersdorf, K. P., The Symbolic Significance of ‘Figurae Scatalogicae’ in Gothic Manuscripts, in Word, Picture and Spectacle, edited by C. Davidson, Kalamazoo 1984, pp. 1-20. Wirth, J., Les marges à drôleries des manuscrits gothiques (1250-1350), avec la col­ laboration d’Isabelle Engammare et des contributions de Andreas Bräm, Herman Braet, Frédéric Elsig, Isabelle Engammare, Adriana Fisch Hartley et Céline Fressat. Geneva 2008.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 173

27-Sep-21 8:09:27 PM

174

bibliography

Zechiel-Eckes, K., Eine neue Arbeitshandschrift des Diakons Florus von Lyon. Der Kommentar des Ambrosius zum CXVIII. Psalm (Cod. Firenze, Bibl. Med. Laur., Plut. XIV. 21), in Révue bénédictine 119 (2009), pp. 336-70. Zetzel, J. E. G., Latin Textual Criticism in Antiquity, New York 1981. ———, Critics, Compilers, and Commentators: An Introduction to Roman Philology, 200 BCE-800 CE, Oxford 2018.

www.torrossa.com - For non-commercial use by authorised users only. License restrictions apply. Troncarelli - Magee.indb 174

27-Sep-21 8:09:27 PM