It is rare honour for a musicologist to have a work published in Corpus Christianorum, a series launched shortly after the Second World War by the Benedictine scholar Father Eligius Dekkers with the aim of providing scholarly editions of early Christian and medieval texts. James Grier has broken new ground by supplying the first volumes to include music. His chosen subject, Adémar of Chabannes (989–1034), has previously featured in the series as the author of a three-volume history of the Franks,Footnote ¹ which was written in part at his adopted abbey of St-Martial of Limoges, the institution that saw the defining moment in his career: the failed inauguration of a new apostolic liturgy for the abbey's patron saint on 3 August 1029. The campaign to promote St Martial to the status of an apostle has been described several times; Grier has spent over two decades immersing himself in the associated manuscripts with music. These volumes represent the culmination of his own single-minded pursuit of what he has elsewhere termed ‘the most stupendous and successful scam of the entire Middle Ages’,Footnote ² gathering together for the first time the ritual component of Adémar's output, which is described on the cover as his opera liturgica et poetica.

The implications of the title become clear in the Introduction. The edition is described initially as a collection of items whose text and music were both copied by Adémar, in which dual capacity he recorded music and texts for the Mass and Offices of St Martial and his companions (i.e. Austriclinian, Justinian and Valeria), as well as items for the Office for St Cybard (patron of his home monastery in Angoulême), a Sequentiary, a Tonary and a handful of other items. Adémar did not compose all that he copied: the opera are the fruits of his skills as a liturgist, poet, musician and scribe, presented as a music and text edition (vol. 245) and a separate text only edition (vol. 245A) framed by an extensive Introduction and Commentary. The double format of this edition with separate accompanying apparatus allows the music to be retained as an integral part of Adémar's enterprise (vol. 245) and the texts to be consulted in a familiar and accessible format by medievalists without musical expertise (vol. 245A).

The main body of the dual edition follows a parallel structure that to a certain extent mirrors the organisation of materials in the original manuscripts by genre:

1. I Tropes of the Mass for Saints Martial, Austriclinian and Justinian

2. II Office chants for the Feast of Saint Martial

3. III Proses

4. IV Song about Saint Martial

5. V Alleluias

6. VI–VIII Office chants for the Feasts of Saints Valeria, Austriclinian and Cybard

7. IX A Sequentiary

8. X Texts for the Feast of St Cybard (text edition only)

Such an arrangement allows those interested in reconstructing liturgy to do so with less difficulty than might be imagined. To piece together the apostolic liturgy for St Martial it is necessary simply to turn to section I.3 for Proper Mass chants and their accompanying tropes for the Feast of St Martial, as well as the troped Gloria. Office chants are located in section II, where antiphons, responsories and versicles from First through to Second Vespers are written out in full, including the psalms. Readings and proper prayers for both Mass and Offices may be found in Appendix F.

A second and simpler way to understand the contents of these volumes is as an edition of Adémar's liturgical dossier in support of the apostolic campaign. Although a useful description to keep in mind, the Sequentiary and Tonary are not wholly dedicated to this purpose, providing sequence melodies across the liturgical year and tonal formulae associated with chants in each of the eight modes respectively. The Office for the Feast of St Cybard is also different in kind insofar as it was copied seemingly after Adémar's return to Angoulême in August 1029, whence he fled to continue his campaign. It is included here as it was compiled and copied by Adémar, written out alongside an apostolic Office for St Martial, and promotes the patron saint of the abbey in which Adémar was residing at the time. Matters become clearer when a third defining feature of the edition is understood, namely the focus on a single manuscript, namely Paris, Bibliothèque nationale de France, MS Latin (hereafter Pa) 909.Footnote ³ This manuscript is familiar to scholars of medieval music as one of the Tropers that passed through the hands of Bernard Itier, librarian at the abbey of St Martial of Limoges c. 1200. Its first layer has customarily been assigned to St Martin of Limoges on the basis of the expansion in the number tropes for this saint compared to the earlier Tropers Pa 1120 and 1121. However, some caution is necessary, for while Limoges looms large in music history due to the survival of its manuscripts, its institutions have yet to be the subject of systematic palaeographical study. The large number of tropes for St Martin in Pa 909 is also not an infallible guide to provenance: a similar number of tropes for St Martin is found in Pa 1119, which is thought to have been written for St Martial. The second layer of Pa 909 contains the fullest record of the apostolic liturgy as compiled, copied and notated by Adémar. It is this layer that is the primary focus of this edition; indeed, items whose text and music were seemingly copied by Adémar and that promote Martial as an apostle, but which were not copied in Pa 909, are included only in the appendices of Grier's edition. Such a description applies to the hymn and responsory verses for St Martial transmitted in Pa 1978, a manuscript transmitting portions of notated Offices for St Cybard and St Martial on its endleaves, which were according to all appearances copied by Adémar at St Cybard of Angoulême.

Certain puzzling aspects of the edition arise from tensions between its three defining axes; that is, between Adémar's involvement as copyist, the apostolic liturgy for St Martial, and the second layer of Pa 909. Several items listed in the various indexes are not edited in either volume, such as tropes for Saints Paul, Lawrence, Benedict and Andrew, as well as for the Feast of the Assumption and the Feast of Saints Peter and Paul. These are listed and receive brief entries in the Commentary since they were notated by Adémar within the second layer of Pa 909 even though they do not relate directly to the apostolic campaign. It is also not the full transmission of the apostolic liturgy that is at issue, but the creation of a liturgy at a particular moment in time. Adémar's autograph is preferred (except on the rare occasions when it is demonstrably in error), followed by readings in manuscripts in which his autograph appears but the entry is not in his hand, then readings from manuscripts that Adémar might have consulted. Further down the priority list come readings from manuscripts whose scribes may have consulted his autographs, followed finally by the broader tradition as recorded in Aquitanian manuscripts.

It is instructive to have Adémar's opera liturgica et poetica collected between four covers, although it has to be remembered that only a small percentage of the material is attributable to him. The strongest identification of Adémar's liturgical composition has been made by John Emerson, who demonstrated that almost all of the liturgical formulae in the Offices for Austriclinian and Valeria were taken from Adémar's sermons and that the melodies are not found elsewhere.Footnote ⁴ Other attributions of liturgical compositions to Adémar are made on two, often related, grounds, namely unequivocal endorsement of Martial's apostolic status and first appearance in a manuscript associated with him. The first raises the question of how far others might have been involved in the promotion of Martial's apostolicity. Evidently there was wider support for the campaign, but the tendency is to assume that Adémar was responsible for the creation of all new apostolic formulae. Taken together, this amounts to a set of Proper chants for the Mass, Proper tropes for the Introit, Offertory and Communion, Sequences and Gloria tropes, and several items for the Office. That certain items first appear in manuscripts connected with Adémar is a more tenuous ground for attribution; for example, the case for his authorship of the melodies for five unique sequence texts in the Proser of Pa 1121 is relatively weak.Footnote ⁵ Here, as elsewhere, the possibility of earlier lost sources is not entertained, no decisive stylistic traits are identified, and the strength of the connection to Adémar via the earliest manuscript witness may be brought into question by a re-examination of the palaeographic evidence (about which more below).

Even in seemingly strong cases for attribution to Adémar, doubts may be raised about the extent of the claims made for his authorship. The prose Nobis annua (III.2.A), which was written for the Dedication of the Cathedral of St Stephen in Limoges, first appears in full in the second layer of Pa 909 and refers to Martial as an apostle. Grier claims only that the text is by Adémar, noting that the melody is the same as that used for the more commonly found sequence text for St Stephen, Celsa polorum (IXA.26.C). What he does not mention is that portions of the text of Nobis annua are found within the earlier sequence text Celsa polorum.Footnote ⁶ Moreover, the sections quoted in Adémar's text appear independently as a prosula for Celsa polorum within the Sequentiaries of Pa 1121 and Pa 909 as reproduced in the first volume of the edition (pp. 600–2). The simplest explanation is that Adémar drew on an earlier text for St Stephen when composing a replacement emphasising the apostolicity of St Martial. It is unclear whether he worked from the full text of Celsa polorum or the prosula; what may be said with some assurance is that the whole of the sequence text Nobis annua cannot be attributed to Adémar.

A similar situation pertains for Apostolorum gloriosa, a sequence text emphasising the apostolic status of St Martial that first appears in the second layer of Pa 909. The first clue as to an earlier version is again provided by unexplained capital letters appearing within verses in Grier's edition.Footnote ⁷ Cross-checking reveals that the text at these points corresponds to a prosula for this melody that appears in the Sequentiary of Pa 909 and is edited in the first volume (p. 545). Since the sequence melody is apparently unique, there is no easy way of tracing another prose text that might have hosted this prosula. Given the model provided by Nobis annua above and in view of the fact that the prosula text again does not mention Martial, is it most likely the Adémar did not compose the whole of the prose text of Apostolorum gloriosa.

Adémar's skill as a compiler may be admired in other cases where he seems to have been responsible for at least partial composition. Both editions open with tropes for the Introit for the Feast of St Martial. As reproduced below, the principles of the edition are followed in giving the trope text in lower-case and the host chant text in upper-case letters.

Concerning St Martial, Apostle of Aquitaine

1. 1. People devoted to God, now receive our song.

   Assuredly we praise the man about whom the sacred words profess:

   [Introit] GOD TESTED HIM AND KNEW HIS HEART.

2. 2. That man is Martial, pre-eminent alumnus of the Lord –

   HE KNEW HIS WAYS –

3. 3. Whom God himself called from the stock of Israel to be a leader.

   HE LED HIM IN THE WAY OF EVERLASTING [LIFE],

4. 4. He sent him, an illustrious man of apostolic standing, into the field,

   AND HIS DOMINION WAS EXCEEDINGLY STRENGTHENED.

   [Psalm] O Lord, thou hast tried me and known me;

   Thou hast known my sitting down and my rising up.

The entire Introit trope (or ‘trope complex’) comprises ten trope sets of which the above is the first. It opens, following established practice, with the newest items; in particular, the second to fourth elements (nos. 2–4) in this first set are found only in Pa 909 and one later source, which, following Grier's line of reasoning, makes it probable that they were composed by Adémar. The first element (no. 1) is found in earlier sources but its deployment at the head of this complex, presumably by Adémar, is highly suggestive. As noted by Ritva Jacobsson, the hexameter couplet echoes the opening line of Vergil's Aeneid by introducing the man of whom the poet sings. She also records that ‘devota deo’ is shared with Prudentius' Contra Symmachum (II:1130), where the reference is to Rome herself as ‘dedicated to God’.Footnote ⁸ Such lofty allusions might have influenced what appears to have been Adémar's decision to place this element at the head of a first set completed by what seem to be his own hexameter lines. As it happens, this is obscured in the literary edition by an apparatus that restricts itself to references only to biblical and other textual sources known to have been used by Adémar.

Pragmatic decisions also seem to have influenced the design of the music edition, in which no attempt was made to articulate the hexameters by starting each line afresh. The blank space that would have resulted may have been prohibitive for an already expansive edition, but what is more immediately appreciable in comparable editions is the relation between melodic design and poetic structure. It is hard, for instance, to gather from the continuous layout adopted the routine marking of both caesura and line ending by melodic cadences on the final or co-final in the first three hexameter lines of the first tropes set for the Introit for St Martial. It is similarly difficult to identify the similar melodic shapes followed in setting the last two hexameter lines of this trope set. The implication of this continuous presentation is that in favourable circumstances the entirety of the items presented on the page would have been performed. Whether an Introit for St Martial lasting circa twenty-five minutes is indeed to be imagined or only one to two sets were sung (the precise choice varying from year to year) is an open question. Some guidance on this point would have helped in attempting to understand the implications for performance of the material contained in the edition. Similar difficulties emerge in attempting to imagine the performance practices of the remaining tropes, proses, prosulae, antiphons, responsories, invitatories, versus and other genres. A brief introduction to these might have extended an already lengthy work, but would have eased the task of non-liturgists consulting these volumes and provided more direction in imagining how the sung items would have worked in performance.

The wider question that is raised by this Introit trope set is the relation of this edition to performance. The decision to reconstruct in full the recitation of psalm verses is indicative of a desire to represent the full sung context for Adémar's opera liturgica. Performers and those reading through the music to themselves will no doubt appreciate the extra effort expended in reconstructing psalm tones, even if those reading the text edition alone may wonder whether printing the texts of psalms and canticles from fifteenth-century St Martial Psalters adds much in terms of imagining what might have been sung four centuries earlier. Providing cues for those interested to look up relevant psalm texts would have been a less controversial option, especially given that the choice of psalms in the Offices is often based on later medieval sources, as explained in the Introduction. It might also have been useful to draw a distinction in kind between the edition with music, where full reconstructions are useful tools in appreciating the totality of a performance, and the literary edition, where interested readers can more easily refer outwards to other texts. A more economical solution would have been for readers of the edition with text alone to have been referred to the edition with music for the full psalm texts and other information about performance.

The presentation of psalms and other multipartite forms in the edition with music is not without its own difficulties. Psalm tones are often divided up in such a way that the two-part structure is obscured by being spread over different lines and pages. Continuous layouts also make it difficult to grasp the five-part structure of the invitatories, the repetitive structure of proses, the strophic design of the versus, and the respond structure of Alleluias. The introduction of a hierarchy of bar-lines (single, double and final) is relatively uncontentious given the adoption of a similar scheme in Solesmes editions among others and the need for sectional clarification within continuous layout. The use of slurs to represent neumes is also understandable, although some will prefer proximity through spacing as a less intrusive way of achieving the same end. Potentially more irritating is the fact that the B♭ signs routinely bump into the slurs above the stemless note-heads.

The most unusual feature of the layout is the listing of melodic variants as notated snippets at the right-hand foot of openings, a practice that follows Grier's own recommendation in his widely read The Critical Editing of Music.Footnote ⁹ Where there are only a few variants, the listing is relatively unobtrusive and simple to use; where there are multiple variants, they take up most if not all of the right-hand side of an opening. The problem becomes particularly acute in the Sequentiary, where pressures of space led to the use of both notated snippets and written comments in text boxes aligned with the passage to which they refer. The potential advantage of this system is that anyone singing through the material can see alternatives at a glance and substitute them more readily than if they had to consult an apparatus presented in a highly abbreviated format. The disadvantage is that the extensive apparatus can seem overwrought for an edition primarily based on a single source to which hardly any corrections are made. It would almost certainly have been easier on the eye to list variants, many of which are matters of liquescence, in textual shorthand either in the commentary or at the foot of the relevant pages. The importance placed on the second layer of Pa 909 in the definition of the scope of the edition might also have been taken as a good reason for assuming that readers would also prefer versions in this manuscript in all but the most exceptional of cases.

Simplification might have been pursued in other ways. A double edition presented in two volumes with nine different main sections (ten in the edition with text alone), an appendix for the Sequentiary and five distinct appendices thereafter (seven in the edition with text alone) is difficult to navigate even before coming to terms with various subsections, the different apparatus for the two editions, and the six different indexes as well as an Onomasticon. It is certainly helpful in attempting to understand the full implications of the material to have a copy of Grier's The Musical World to hand, where almost exactly the same numbering system is used and further detail about individual items is often to be found.Footnote ¹⁰ Even so, focusing on only one of the defining axes of the edition would have considerably lightened the editorial task. It would have been a simpler exercise, for example, to have concentrated on the apostolic liturgies, thus radically cutting down the Sequentiary and Alleluia incipits, excising the Tonary, and expunging the pre-apostolic Offices of Martial and Valeria. Equally, the two separate editions might have been rolled into one with a more exclusive focus on Paris 909 as a best source. Finally, full reconstruction might have been put aside, replacing fully copied-out psalms with first verses and cross-references to modern sources for further verses.

The point of highlighting potentially simpler editorial approaches is not to imply that Grier has fallen short of his own goals, but to show that he has favoured comprehensive solutions at every turn. The Introduction is a case in point. Emphasis is placed on explication of the manuscript tradition, concentrating in turn on evidence for the identification of Adémar's music hand, a reconstruction of Adémar's musical activities through manuscript witnesses, a description of individual manuscripts used beginning with Adémar's autographs, and a discussion of editorial principles focusing in turn on manuscript relations, techniques used in reconstructing the apostolic liturgy, and editorial presentation (including an extended consideration of the treatment of B♭s). The Commentary, which occupies most of the first volume, is similarly extensive, providing a full account of sources consulted and decisions made in reconstructing pitch levels of individual chants. The reader is assured that Grier has left no stone unturned.

The level of detail entered into at every point ensures that scholars will profitably read these volumes for decades to come. It also means that no review of moderate length can adequately address every aspect of these volumes; what has been done here is to concentrate on fundamental matters of definition. Even so, to conclude at this point with a summary appreciation of these volumes would be to make two serious omissions. The first is the need to acknowledge the extraordinary dedication with which this edition has been prepared. While Adémar's story makes a great tale, it has also tended to obscure the skill with which he pursued his goal. With the benefit of this edition, it is now possible to appreciate how he had to master all genres of liturgical composition in order to accomplish his ends. The difficulty facing the modern editor is the need to comprehend a similarly wide range of genres, a skill equally, if not at times more, to be admired. To give a single example, in order to reconstruct the troped Gloria Rex apostolorum Deus, it was necessary to work out which Gloria melody was used, which Gloria trope using this melody for its closing section was replaced on this occasion by a new trope text, which versions of these melodies and texts most closely correspond to the versions in Pa 909, and on what pitch level to begin the reconstruction. While it remains relatively simple for anyone to express a preference one way or another about editorial decisions made in presenting individual items, the sheer scale and complexity of the task undertaken deserves recognition. It is also no small matter to observe that both text and music are presented to a high level of accuracy throughout the two volumes. The whole is an extraordinary testament to Grier's extended fascination with the work of a figure who was similarly complex, idiosyncratic and highly inventive.

The second area that requires separate consideration is identification of Adémar as a notator. As Grier himself notes, his previously published comments on Adémar's music hand have been ‘introductory’ and a full palaeographic discussion is forthcoming in the series Corpus Christianorum Autographa Medii Aevi. The only musicologist previously to provide palaeographic detail in discussing Adémar's notational hand is Michel Huglo, who briefly discussed three manuscripts containing text partially copied by Adémar that might contain his neumes (Pa 1121, Pa 909 and the closing leaves of Pa 1978). No definitive features of Adémar's notational hand were identified, but Huglo did mention in passing three signs that could form the basis of a future study (the custos, the quilisma and the pes stratus). Grier extends Huglo's observations, as well as more or less tentative and circumscribed claims made by others, by asserting that Adémar notated the bulk of the former two manuscripts. The most detailed criteria for these claims published to date are found in the Introduction to this edition, where three distinctive features of Adémar's neumes are listed: the custos, the bowed angle of strokes in the clivis, and the slanted orientation of the neumes arranged in vertical descent. These three distinctive features are not in themselves sufficient for identifying Adémar's notational script. As noted by Grier, the custos assumes a different shape in Pa 1121 (a horizontal dash) from that in Pa 909 (where the dash is preceded by a rising left to right diagonal). Given that there are sections of Pa 909 where a custos in the latter shape is used alongside neumes not attributed to Adémar, the form of the custos cannot be taken as a sole criterion for identifying his notational hand.Footnote ¹¹ The two neumes mentioned also cannot be claimed as decisive criteria for identifying Adémar's hand until a wider ranging study of notation in these and contemporary Aquitanian manuscripts has shown that these particular features are unique to his script.

Grier's claim to have identified Adémar's notational hand is therefore open to scrutiny. There are at present two ways into defining Adémar's neumes: first, by proximate association with his text hand, and in particular the apostolic liturgies as compiled in Pa 909 and Pa 1978; second, by means of a colophon added to the end of the Sequentiary in Pa 1121. To begin with the former, Adémar has been identified as the scribe of both the second layer of Pa 909 and the Offices added to the closing folios of Pa 1978 through palaeographic comparison of these scripts with other manuscripts known to have been copied and in some cases signed by him.Footnote ¹² The issue at hand is whether the same hand also added the neumes. This is not the place to begin the sort of detailed assessment required to prove the case positively, but what may be said here is that there is no clear evidence that different notators added neumes to the second layer of Pa 909 and the endleaves of Pa 1978. To refer only to Huglo's touchstones, the custos assumes the same shape in both and the quilisma is written at a similarly sloping angle; the pes stratus is not clearly visible in published facsimiles of Pa 1978, but in the second layer of Pa 909 its diagonal slant is particularly noticeable. Grier's claim that Adémar also added neumes to the first layer of Pa 909 is harder to assess using the same criteria: the clearest difference lies in the pes stratus, which tends to be written with an upright initial descending stroke in the first layer, thus routinely giving the first bow a deeper appearance than in the second layer where the first stroke is often curved. With this said, isolated examples of the tendency associated with the first layer can be found in the second layer and vice versa. It would seem prudent at this point to reserve judgement until the advent of further detailed study, especially since there are clear examples of both text and neumes added by different hands within what is signalled as an undifferentiated first layer of the manuscript by Grier.Footnote ¹³

The colophon added to the end of the Sequentiary in Pa 1121 is open to different interpretations. It is simplest to note here that the closing words (Ademari … qui rite hunc biblum notavit) indicate that Adémar notated something and that the most straightforward interpretation is that he added neumes to the Sequentiary. Any attempt to follow this up through study of neumatic forms soon runs into difficulties as the general appearance of the neumes differs markedly depending on whether the beginning or end of the Sequentiary is taken as the starting point. What may be traced by reading through the folios of the Sequentiary in Pa 1121 is an overall development through the two gatherings, tending towards a neater and more compact style with more sharply angled forms. In addition, significant changes in neumatic formation are introduced from the beginning of the second gathering.Footnote ¹⁴ It is possible that a single notator was experimenting with how best to meet the demands of notating a Sequentiary, improving the layout and increasing the precision with which pitches were signalled by the individual neume shapes. It also remains possible that a new notator began work at the second gathering. Before reaching any conclusions in this area, further study of this fascinating notation is required.

Perhaps the safest claim that can be made at this point is that the neumes in the second layer of Pa 909 and the Offices added to Pa 1978 have the strongest case to be considered the work of Adémar. Comparison of these neumes with those in the Sequentiary of Pa 1121 throws up several points of difference; in particular, the neumes in Pa 909 are written at more of a sloping axis, with the tractulus (the dash representing a single note) routinely written at a rising angle, whether or not the melody is ascending. There are also marked differences in the manner of writing individual neumes; for example, clives signs in Pa 909 include even more sharply angled signs alongside the more rounded ones, and the final tick at the end of the form now assumes the same length and angle as the tractulus. Whether or not such differences are sufficient to rule out scribal identification remains to be seen. Grier's claim that the same scribe notated the whole of the first layer of Pa 1121 is similarly difficult to verify. The simplest initial comparison is with the Proser (fols. 196^r–201^v) as many of the same sequence melodies are recorded there, albeit now texted with accompanying prose. In general terms, the notation in the Proser is thinner in the width of its strokes and the vertical strokes in both text and notation are more upright. These general features can be observed in the sharper ascent and acuter descent in the clivis, and the more erect quilisma and liquescent pes. An additional difference in formation can be observed in the flatter execution of the final stroke of the pes stratus.

Significant caveats may therefore be raised about Grier's claim to have identified Adémar's neumes throughout the bulk of Pa 1121 and Pa 909. This is not to say that his claims are wrong; the point is rather that his assertions remain unsubstantiated. It might be that Grier goes on to prove his case in the promised forthcoming palaeographic study. Until he does, the claim that this edition represents the opera liturgica et poetica of Adémar of Chabannes must await further elucidation.