In the summer of 2007, as part of the University of Oxford's Classics Outreach Programme, members of the Classics Faculty put on a performance of ‘Aeneid: The Musical’, the publicity for which announced that ‘Virgil's tale of the founding of the Roman Empire is being brought to life in an accessible way, accompanied by modern songs such as “Bad” by Michael Jackson, “Together” by the Village People and also songs from, Dido, U2 and Robbie Williams, together with classical music from Vivaldi, Stravinsky and Purcell.’ Footnote ¹ Whatever one makes of the choice of songs, the intersection of dramatic staging and the classics has a long heritage: snippets of evidence from the first millennium include a donnish aside by Augustine (‘Few of you know from books, many from theatres, that Aeneas went down to the underworld’ – p. 177, n. 18), and a report that Virgil's own recitation of his Eclogues proved such a hit with his first-century BC audience that extracts were subsequently sung in the theatre by a courtesan, much to the astonishment of none other than Cicero (pp. 175–6). But what of the music that might have accompanied first millennial performances of the classics? Little can be gleaned from early reports of recitation in a variety of different contexts; the first glimpse of what it meant to sing the classics comes with surviving neumatic notations added to manuscripts from AD c. 900 to c. 1200.

Making the scope and nature of surviving notations clear to both Latinists and musicologists inclined to overlook a seemingly esoteric field is one of the laudable aims of Ziolkowski's monograph. The extent of the topic is established at the outset with reference to documentary evidence assembled in two appendices, which introduce and reorder instances of surviving notations of classical verse first collated in studies of the manuscript transmission of classical authors. The story of how the philologist Birger Munk Olsen compiled an inventory of manuscripts transmitting the Latin classics from the ninth to twelfth centuries, and how the codicologist Yves François Riou gathered together information about neumatic notations in a series of publications in the early 1990s as a way of gaining insight into provenance has been told several times. The results are condensed by Ziolkowski into a single sentence: ‘[N]eumes are extant for nearly two dozen of Horace's Odes as well as for one section of the Carmen saeculare and for parts of two Epodes; for more than a dozen in six books of Lucan's De bello civili (alternatively, and less properly, known as the Pharsalia); for eight in seven books of Statius's Thebaid and one in his Achilleid; for four in two of Terence's comedies; and for two in two of Vergil's Eclogues, two in the Georgics, and more than two dozen in ten books of the Aeneid’ (p. 5).

Venturing into relatively unfamiliar waters for a Latinist, especially one as eminent as the current Arthur Kingsley Porter Professor of Medieval Latin at Harvard University, Jan Ziolkowski is quick to disavow any new-found musicological expertise. Deference over interpretation of neumatic detail, which is a necessity for anyone wishing to comment on musical content since hardly any of the recorded melodies can be reconstructed with any certainty, is paid to Silvia Wälli's work on the surviving Horace notations and Gundela Bobeth's forthcoming study of the remaining notations of classical verse.Footnote ² In seeking to approach an interdisciplinary field through his own set of interests, emphasis is placed instead on what the surviving notations reveal about the reception of the classics, a not altogether surprising move for one who recently published a jointly edited tome running to over one thousand pages entitled The Virgilian Tradition: The First Fifteen Hundred Years.Footnote ³ Yet despite repeated disclaimers, Ziolkowksi is remarkably well informed about neumes and brings a fresh perspective to bear on their interpretation by considering them as part of a wider group of signs used to convey information about text. Drawing on an argument made by Leo Treitler, graphic similarities between neumes, punctuation, prosodic and gloss signs are held to imply a broad functional similarity, the neumes detailing a specific reading of certain prosodic features in what might be termed a heightened mode of expression.Footnote ⁴ A number of intriguing cases are considered within this framework, the most striking of which point to the similar aspect and potential functional overlap between prosodic and neumatic signs.

The first question normally asked about this body of material concerns the context in which the notated classics were performed. While the possibility of admiration of sung classics among an elite circle is mentioned, by far the majority of attention in this study is given to educational use; moreover, the coherence of this set of material is seen to lie in the emphasis placed on this set of authors within early medieval education rather than in any rigid distinction between ‘classical’ and ‘postclassical’ literature. A familiar passage describing the teaching of the most famous magister of the tenth century, Gerbert of Reims (d. 1003), underlines the presence of the relevant set of classical authors within at least one scheme of monastic education (p. 34):

Significant headway is made in tracing symptoms of school use in the neumed manuscripts themselves. Proceeding from the observation that the authors notated were also those most intensively copied from the ninth to twelfth centuries, attention passes to circumstantial evidence that might be brought to bear upon interpretation of the material recorded in the codices, encompassing consideration of the presence of multiple and on occasion vernacular glosses, notated passages in and among commentaries, scansion and construe marks, and a number of occasions on which the neumator can be identified as a corrector, which in itself suggests the work of a magister in charge of both grammatical and musical teaching. In and around assessment of the evidence provided by individual manuscripts, space is also set aside for imagining didactic scenarios for the neuming of the classics, with preference ultimately given to the likelihood that it was mainly schoolmasters or cantors who notated passages in preparation for their lessons.

Having established a strong case that some kind of didactic intent lay behind the addition of notations, possible motivations for neuming only select portions of the classics are considered. Formal possibilities are explored first; namely, that singing and by extension musical notation aided the learning of metre, served to project accent and/or ictus, aided syntactical and semantic comprehension, promoted syllabification and helped pupils to memorise passages. For all the permutations that might be imagined, Ziolkowski sagely concludes that no single thesis can be advanced while musicological research remains at preliminary stage. A couple of suggestions are nevertheless worth pursuing further; in particular, that memorisation through music might have preceded metrical comprehension, and that, in a manner broadly comparable to the pointing of Old English verse, neumes served to mark rhythmical phrasing, providing a frozen record of a private performance in the mind or vocal chords of a reader. Subsequent reflection upon the practice of selecting only excerpts for notation initially entertains the possibility that extensive portions of the classics were sung to the type of simple formulae that would not have required notation; in other words, only melodically unusual and otherwise important passages were neumed. No definitive position is reached on this question, although parallels are drawn with the limited evidence that simple melodic formulae were used in reciting vernacular epics.

In the absence of hard evidence for extended recitation, most attention is paid to the textual qualities of the passages neumed. As observed by several previous scholars, speeches and laments elicit the most notational activity; to mention only the most frequently neumed, Cornelia's bewailing of her ill fortune in Lucan's De bello civili, Argia's lament for Polynices in Statius' Thebaid, Aeneas's vision of Hector and Dido's prayer that she be left alone to grieve. The observation that neumed passages either share certain themes (the macrocosm, the history of Rome, love, etc.) or represent high points in their narrative suggests a broader frame for interpretation: although the point is not drawn out explicitly, what is being observed is a collection of themes comparable to those in contemporary song repertories, whether planctus, computus songs, love songs, or narrative songs. Most pertinently, Ziolkowksi stresses that emotional appeal also played an important role, the pathetic quality of speeches and laments in particular exciting a response. A highly suggestive parallel is drawn with the type of rhetorical exercises undertaken in antique schools, especially the discipline of character description, which typically involved writing women's laments for the dead. Having observed a concentration on expressive delivery in neumed passages, something of a full circle is drawn by exploring in some detail the possibility that at least one of the functions of neumes was to mark out spoken passages at a time when fewer graphic signs and typographical conventions were available to distinguish passages of direct speech.

The teasing out of parallels for notating the classics continues by considering evidence for public recitation of classical verse before the earliest notational records. Several accounts of early theatrical performances of at least portions of classical texts are discussed, although it is underlined that the only private poetry for which singing can be directly attested is Virgil's Eclogues. Recitation within an educational context is also reflected upon through a few striking examples, including Ausonius' (c. 310–95) recommendation to his grandson that he read the lyrics of Horace, Virgilian epic and the comedies of Terence, emphasising in his reading modulation, stress, accent, expressiveness, punctuation and pauses. As to the quality of theatrical or school recitation, caution is exercised given the paucity of both available evidence and previous research into stylised delivery or pronuntiatio. Even so, the relative likelihood of a range of possible modes of delivery is explored, supported by mention of a third, intermediate voice between that used for prose and melody in the writings of theorists such as Aristides Quintilianus, Martianus Capella and Boethius, all of whom relate the intermediate voice to poetic recitation. For all the hints of musical elements in classical and late antique poetical recitation, a strict dividing line is drawn between such evidence and surviving medieval notations, the latter regarded as the technical and cultural product of a wider Carolingian renovatio. More influential models are traced in both sacred and secular laments, whether in ecclesiastical traditions in which certain books of the Bible were thought to have been originally composed in hexameters, and laments of Mary and Rachel were held in particularly high regard, or in accounts of female laments sung beyond the reaches, but not the hearing, of the Church.

In view of the associations drawn throughout the study, it is perhaps not surprising that the decline of neumed classics is traced to the waning of an ecclesiastical educational culture. While a host of factors is considered, the key consideration in the effective conclusion of the tradition of neuming classical verse by the thirteenth century is held to be changes in schooling. Around this time the disentangling of literacy and liturgy, with all that the coupling had entailed for ritualised reading and singing, accompanied a shift to more private forms of devotion and reading. A second key factor is held to be a growth in new compositions from the twelfth century onwards, not only in the field of Latin rhythmi composed according to regularised principles of accent and syllable count, but also in the vernacular. In both fields, classical themes were taken up within new forms of expression, such as Dido's predominantly rhythmic Latin lai in the Carmina Burana and several vernacular lais on the same theme. In addition, the twelfth century also saw exploration of classical themes in the new romans d'antiquité in Old French with their equivalents in Middle High German, whose heroic tales updated classics such as the Aeneid and the Thebaid.

In covering a wealth of material with an expertise that lies beyond even the most ambitiously literary of medieval musicologists, Ziolkowski's monograph comes as a welcome intervention in a field recently dominated by highly detailed manuscript research. His panoptic vision serves to bring out similarities and differences over the longue durée, while, as a voluminous reader with an ability to turn a phrase, he is also particularly well placed to summarise multiple threads of evidence and come to an assessment of the historical importance of a set of neumations that have for a long time been regarded as something of a backwater in the history of medieval music. There are, of course, details over which specialists might quibble. Ziolkowski takes a conservative stance in dating the practice of singing the classics to the late tenth century onwards, yet whether the onus should fall so heavily on the date of surviving notations is by no means certain. He is aware of limitations imposed by the surviving notational evidence insofar as he includes in Appendix I unnotated extracts that appear in collections of material intended to be sung, his principle point of reference here being his own work into and convictions about the compilation of the Cambridge Songs (Cambridge, University Library Gg. V. 35).Footnote ⁵ Extending the same principle to earlier collections, a strong case could be made for entries in a verse collection copied in the region of Paris in the middle third of the ninth century and known to have been notated by a magister at Laon Cathedral in the early tenth century (Berne, Burgerbibliothek 455). All the items in the verse collection (hymns, late antique metra, rhythmi) are suited to musical performance, among which are the opening thirty-three lines of Virgil's first eclogue (Tityre tu patule) copied out in three-line stanzas.Footnote ⁶ Adding this passage would not only expand the number of classical texts believed to have been sung, but also put the earliest date of at least some version of this practice back to the middle of the ninth century.

For an author such as Virgil, whose transmission is attested from late antiquity onwards, sung performance of select passages might also be considered at an even earlier date. The listing of Lucan, Statius and Virgil in Alcuin's poem on York of the early 780s, and the presence of the same plus Terence and Juvenal in the catalogue of Charlemagne's court library of c. 790 points to an institutional investment in such authors by the late eighth century. Material evidence for singing is, of course, harder to come by, but there are numerous references to singing alongside copious classical allusions in the poetry emanating from the scholars gathered around Charlemagne and his court. We might speak of an ideal of learning based around reading and singing canonic literature fostered in the formative years of the Carolingian renaissance, the working out of which required the alignment of classical scholarship with Christian education at larger monasteries and cathedral schools in the following centuries. The Berne manuscript might also point to a more fluid early practice of singing the classics, possibly attached to a smaller and slightly different selection of texts than represented by the later notations, which, as far as one can discern from the limited palaeographic information provided, cluster around Ottonian abbeys and cathedral schools from the mid-tenth century onwards.

It would be remiss to end without considering in further detail the decisions that inform the compilation of the two Appendices. One of the reasons that the corpus of notated classical verse has only slowly come to wider attention has been the difficulty in gaining access to and making sense of Riou's formidably abbreviated tables. In presenting the material afresh, emphasis is placed squarely upon the texts, with author and passage forming the main categories and basic manuscript information (including date, origin, provenance and type of neumatic notation) repeated for each extract. Although this leads to a fair amount of repetition, the task of surveying the surviving musical evidence for any particular passage is considerably eased. The second Appendix, which lists a subset of notated speeches, is also tailored towards a particular audience, allowing those already familiar with the contents of the listed texts to come to their own judgement about the reasons for the distribution of musical notation. A division in audience expectation is further marked in the inclusion of material subsidiary to Riou's earlier list in Appendix I. Musicological expertise is most explicitly called for in considering the potential relation of marginal notations to main texts: the relation of one listed Horace case was comprehensively dealt with in Silvia Wälli's edition; judgement on several Virgil cases is best left to Gundela Bobeth in her forthcoming edition and commentary on all the surviving neumed classical verse apart from Horace. Songs transmitted in theory treatises also demand further specialist attention insofar as they are separated off into the prose passages that preface the lists of neumed passages. Such a division of information is telling, confirming that Ziolkowski's primary interest lies in cultural practices (reading and singing the classics) as opposed to a comprehensive reception history of sung classical poetry up c. 1200, which would need to concern itself not only with further research into specific personalities and places, but also with the presence of lines of classical poetry in other song repertories.

Ziolkowski's boldest decision in compiling the appendices is to include preliminary information about neumed passages of Latin verse by authors writing up to the period of the earliest musical notations. In a characteristically generous gesture, he proceeds by gathering together available evidence and stresses the provisional nature of the lists. While it would be possible considerably to extend the number of known notated verses by rummaging through unpublished working lists and a few further publications that mention neumes, musicologists have tended to be ponderous in releasing the fruits of their labours in this area, in part because any listing of notated passages remains highly provisional in the absence of systematic recording of neumes in manuscript catalogues and text editions, and in part because boundaries become hard to define with later texts. To cite a single case, neumes are mentioned in Carl Springer's handlist of Sedulius manuscripts, but the information is at times incorrect, and when notations were added to verses that were later adapted for use as hymns, such as A solis ortus cardine, it becomes difficult to pick apart overlapping educational and liturgical usage.Footnote ⁷

Perhaps the best way of promoting the type of scholarly exchange envisioned by Ziolkowski would be by means of an online database, which would allow information to circulate and be updated without the need to commit to the finality of print. Such would certainly be useful given the piecemeal information about musical notation in existing indexes such as Initia carminum Latinorum saeculo undecimo antiquiorum or the database In principio.Footnote ⁸ More useful still would be a search facility allowing entries by author, passage, verse form, type of notation, date of manuscript, type of manuscript, origin, provenance and so forth, as well as information about published facsimiles, transcriptions and commentaries. Such is the business of research proposals; Ziolkowski's demonstration of what can be done when a previously obscure topic is opened up to those with differing expertise is a more eloquent expression of the need to develop more flexible research tools in this area.

In a field where pioneers are needed, this monograph stands as a clarion call to musicologists to continue exercising their patient detective work in seemingly unpromising fields in the assurance that others are taking note. With that in mind, tribute should be paid here to the work of the late Silvia Wälli, who first showed what could be done from a systematic musicological perspective with such apparently unpromising material. It remains, as Jan Ziolkowski has urged, for musicologists to complete work begun in this area so that all may enjoy resurrecting the classics through song.