Leo Treitler concluded his seminal study ‘Reading and Singing: On the Genesis of Occidental Music Writing’, published first in these pages thirty-five years ago, with the obiter dicta that understanding the origin of music writing ‘is a question of semiotics, not alone of paleography’.Footnote ¹ While not denying that music writing is a system of signs, Susan Rankin’s important book argues that the sudden appearance ‘within a period of probably not more than fifty years (and probably rather less)’ (p. 328) of varieties of music script in the ninth century can best be understood from the perspective of palaeography. During the course of the ninth century music books and fragments of music came to be accompanied by supralinear notations. By the end of the century music notation was commonplace and had developed its own broad conventions. But was the ‘invention of musical notation’ a paradigm shift? In an early example of the careful and judicious reasoning that characterises her investigation, Rankin answers that it was not (pp. 77–84). Memory was at the heart of music production, obviously so when music texts did not come with signs, but even so when music texts did come equipped with signs. Memory now worked in tandem with signs, signs that were written with the same quill and ink as the alphabetic signs they accompanied and thus the subject matter of palaeography.

Rankin’s new narrative succeeds on several levels. It offers a panoramic and nuanced view of Carolingian music notation that deftly synthesises decades of musicological research while opening up original and bold perspectives on the development and cultural significance of ninth-century notation. The series in which the book is published suggests the inspiration for its fresh approach: Rankin’s investigation of music notation is nested in a much larger framework, the persistent and programmatic Carolingian emphasis on writing, which restores agency to the individual music scribe. Writing Sounds is also an archaeological investigation that starts with well-developed music scripts of the late ninth century to try to uncover earlier stages in the prehistory of notation. This could be risky business, but for the author’s careful methodology and keen understanding of the cultural context of music production in the Carolingian age. Readers expecting a schematic, genealogical history of the development of early music script will be disabused. History can be messy, a constant Rankin keeps in mind when she refers to the development of music script as ‘chaotic’ over a fifty-year period of rapid innovation and experimentation.

Running throughout Writing Sounds is an implicit comparison of the development of music script with the development of minuscule script. Both were predicated on the correct preparation and performance of the word of God (pp. 340–53). The gradual, uncoordinated adoption of minuscule script in Carolingian lands enhanced legibility without obliterating different letter forms and systems of abbreviations and contractions. By the time Carolingian music script was fully developed at the end of the century, it too, within a generally coherent system of recording with ink marks on parchment, did so with variety born of experimentation and innovation. The reciprocal relationship between the two types of script is no better shown than when music texts were copied in a new way with words either broken up by significant spaces or bunched together to run parallel with the appropriate musical notation inscribed above them. Here, in one of many helpful comparisons, Rankin invites her readers to consider the differences between a Fulda manuscript of c.830 and a Laon manuscript from the last quarter of the ninth century (pp. 65–72). In the earlier gradual, there is no relationship between its text and its music script in terms of the manuscript’s layout. A half century or so later, the individual text syllables of the Laon gradual were laid out to coordinate with its accompanying neumic notation.

Another major theme running through the book is the reminder that readers should eschew teleology, the facile assumption that Carolingian music script was the first, albeit primitive, stage on the way to modern notation. In Rankin’s telling the rapid development of music scripts in the ninth century was no simple ‘changeover’ from oral to written music. A person unfamiliar with chant would find it impossible to make song guided only by the neumes on the page. Proper interpretation of music signs required first knowing a song’s melody. The visual information provided by the notations guided the singer’s recall of the song’s melody. The two, memory and sign, worked together to produce correct song. This makes sense. Monks, nuns and cathedral canons would have listened to and sung melodies from their earliest years and would have had a song’s basic tune committed to memory. To sing it correctly, or differently on some occasions, required tutelage from their communities’ specialists who, armed with a neumed song book, could show the way.

Rankin also makes sense when she accounts for the evolution of music scripts. Earlier attempts to construct a stemma as one does for literary texts to show relationships among the various music scripts are wrong-headed. Music scripts are scripts and the project to delineate relationships depends on palaeography. Writing Sounds is firmly grounded in Bernhard Bischoff’s roster of ninth-century manuscripts.Footnote ² Not only did Bischoff provide dates and locations for manuscripts, he was also particularly attentive to record the presence of musical notation in all kinds of books. (An appendix, pp. 363–9, of ‘Manuscripts with Notations Written in the Ninth Century’ lists 120 items.) This is the research base Rankin plumbs to present a revised taxonomy of six Carolingian music scripts and to offer an account of their evolution. Her methodology, applying the linguistic category of cognates from historical linguistics to musical notes, is the trowel and brush she wields to decipher the archaeology of ninth-century categories of music scripts. Setting aside regional groupings, since conventionally defined notations are more alike than different, she interrogates individual graphs to find meaning and connections: ‘[I]f signs set in similar contexts are written with similar forms, then those forms may be cognate, that is, there may be a historical relation between them’ (pp. 185–6). In addition to scrutinising the graphic forms of notes across the spectrum of manuscript witnesses, the investigation also considered how the forms were combined and entered in the interlinear space above (or, in the case of Nonantolan script, sometimes below) their accompanying texts. Applying these criteria to the research base resulted in identification of six different types of music scripts by 900: Palaeofrankish, Frankish, Breton, Lotharingian, Aquitanian and Nonantolan. Each of these is described in detail to establish a context for the development of music script. That context was highly dynamic and was driven by music scribes familiar with different types of scripts who adapted them to their own needs. The scripts, far from being created independently in musical silos, as it were, show a very high degree of interrelatedness grounded in a common sign repertory. The breakthrough to the very idea of representing music through graphic forms came, as Rankin tells it, when musicians adopted the accent marks (acute, grave, circumflex) grammarians used to guide the speaking of sound and applied them, along with dots and dashes, to the task of singing sound. This happened first with the Palaeofrankish script type, where the new script indicated not notes or the movement of sound between notes, but simply change in the movement of sound. Other scribes adapted these basic forms and the technique of using interlinear spaces for their work to create new scripts. The Frankish script type marked an ‘enormous revision’ (p. 319) of the Palaeofrankish when it used the signs to indicate notes rather than movement. Musicians were making the grammarians’ marks their own and using them for new purposes. These initial steps unleashed fifty years of ‘exploration, experimentation, exploitation’ (p. 358) and, one might add, cross-pollination that led to the mature forms of the scripts c. 900 and the adaptation of text layout in manuscripts to accommodate them.

But what prompted the ‘invention of musical notation’ in Carolingian lands, especially since ‘nothing like it was invented anywhere else’ (p. 318)? Here Rankin opens the lens from the characteristics of individual music notation types to explore broader cultural issues affecting literacy in general and inspiring musical literacy in specific. Carolingian leaders are well known for having seized on literacy as the key element of their audacious effort to reform and correct Frankish society. In his late eighth-century circular letter on the study of literature, Charlemagne chided monks for their poorly expressed sentiments in correspondence received at court. What worried him was that inattention to language would lead to lack of wisdom in understanding Scripture, the bedrock of the Carolingian world-view. It was not sufficient that people should be impressed by the sight of monks; more importantly, they should be edified by their correct reading and singing. Further directives centred on scribes and cantors, who were bidden to copy texts with care and ‘to perform the sounds of the words clearly and elegantly’ (pp. 342–50). Thus correction and precision were in the air when cantors read in the grammarians that accent marks might be of use in controlling sound. One other bit of Carolingian encouragement should be thrown in the hopper when accounting for the new emphasis on music literacy. Charlemagne’s General Instruction was broadcast in 789 to spur a wholesale reform of Carolingian society. Its guidelines were directed at bishops, priests, abbots and monks, the foot-soldiers of the reform. In chapter 70 of the document, religious leaders were told: ‘Let schools be established in every monastery and bishopric for boys to read psalms, notes [notas], chant, computus, grammar, and well corrected catholic books, for often when someone desires to beseech God effectively, they ask poorly because of uncorrected [catholic] books.’Footnote ³ Given the date of the General Instruction, it would be adventuresome to see neumes implicated in the order to teach boys to read cantus. But the General Instruction was copied often and widely disseminated in the ninth century and the idea of ‘reading music’ could have abetted the invention of musical notation.

Writing Sounds comes amply illustrated by thirty-five figures from manuscripts, some of which depict music texts as endleaves or fragments, often in poor condition, suggesting that yesterday’s notations were no longer useful to a new generation of practitioners. Seventy-five examples juxtapose script forms and musical passages from different manuscripts to illustrate similarities, differences and development. These are conveniently for the reader embedded quite near the author’s discussion of the topic they exemplify. Indexes of manuscripts and of chants and songs are helpful, but such a rich and detailed investigation merits more entries than occur in the spare General Index (pp. 398–404), for example, ‘grammar’, ‘language’, ‘liberal arts’, ‘literacy’, ‘manuscript layout’, ‘semiotics’, ‘speech’ and ‘sound’, all topics important to the book’s argument. That is a quibble. Make no mistake, Writing Sounds is a tour de force of scholarship sure to become the foundational study of Carolingian music scripts for generations.