This is not the first book to be dedicated to the offertory chants of the Mass, but it is by far the most comprehensive. In 1939 the Swiss Capuchin Hubert Sidler published Studien zu den alten Offertorien mit ihren Versen, an investigation of the first sixteen Gregorian offertories of the liturgical year (Ad te domine to Domine vivifica) intended as a contribution to the restoration of the melodies and their verses.Footnote ¹ The recent Alle sorgenti dell'offertorium: Canto e rito (Košice, Slovakia, 2008) by Ján Vel'backý, originating in a dissertation at the Pontificio Istituto di Musica Sacra, furnishes an overview of the repertory with instructive tables that compare the chant texts with their biblical sources, as well as pages of musical examples illustrating the ‘toni dei versetti’ of the Roman offertories.Footnote ² Rebecca Maloy's Inside the Offertory: Aspects of Chronology and Transmission (revision of a 2001 University of Cincinnati dissertation) brings our understanding of the offertory chant many leagues forward.

Maloy does not delve into the origins of the offertory as a liturgical act, perhaps because the available evidence has already received sufficient scrutiny. Placing bread and wine on the altar was from early Christian times a simple functional matter. How (and why) it became the meticulous ceremony of the papal Mass described in Ordo Romanus 1 (c. 700) remains imperfectly understood; reference is made to the offertory chant only when the singing is about to conclude. The Hadrianum, an eighth-century papal sacramentary in use at the time Ordo Romanus 1 was compiled, refers to an ‘offertorium’ preceding the bishop's ‘oblatio’, but without giving details.Footnote ³

Inside the Offertory considers two complete but parallel chant repertories, (Old) Roman and Gregorian, that are based on quite different compositional principles. (The reason for retaining the traditional ‘Gregorian’ nomenclature, instead of ‘Romano-Frankish’, for example, will become apparent later.) These very long pieces consist of a respond with one to four melismatic verses.Footnote ⁴ Virtually all of the Gregorian offertories, apart from a few contrafacta, are unique chants, but many Roman offertories make prominent use of two reciting formulae, chosen according to the mode of the piece, as well as multiple repetitions of longer and shorter phrases. (This phenomenon differs from the immediate repetition of text and music – something unique to the Roman and Gregorian offertories and still unexplained.) But even after formulae and repetitions have been taken into account, Roman offertories still have much ‘free’ music. Inside the Offertory expands the frame of reference to embrace ‘cognate’ chants in the Beneventan, Milanese and Old Hispanic repertories, the latter recoverable only in unheighted neumes whose exact pitches cannot be transcribed. Hypotheses of Gallican influence, textual or musical, inevitably encounter the problem that no Gallican ‘offertory’ chants survive, if such a repertory ever existed.

Inside the Offertory is divided into two large sections. The first (chs. 1–5) presents a wide-ranging discussion of the Old Roman and Gregorian repertory, text and music, supplied with references to parallels of Beneventan, Spanish (sacrificia) and Milanese (offerendae) provenance. Chapter 1 evaluates previous research on the offertory, on the relationship between the two chant ‘dialects’, and on the introduction of Roman chant to Francia. Maloy explores the disparate conclusions that have been drawn from more or less the same evidence, adroitly weaving conflicting views together in a manner that demonstrates her superb mastery of the secondary literature. Here and throughout the book, she presents the opinions of others with fairness and balanced judgement.

Chapter 2 surveys the Latin texts of the offertories, raising questions about the possible non-Roman origin of some offertory texts. Readings that apparently depart from the text of the Roman Psalter are tabulated in the associated appendix 1. Chapter 3 compares the music of selected Gregorian and Roman offertories. Chapter 4 explores the question of ‘Italianate’ musical style, mainly in conjunction with Milanese chant, the only other substantial peninsular repertory besides the Roman to have survived. Based on the findings of these chapters, Maloy tackles in chapter 5 the thorny issues of ‘Origin and Chronology’, reiterating previous conclusions and presenting new ones.

The second section (ch. 6 with appendices 2 and 3) provides a description of the sources and an impressively detailed analysis of each of the ninety-four pieces in the core repertory. This draws on the obvious Roman sources along with a select number of diastematic and adiastematic Gregorian sources: graduals and offertorialia (manuscript collections of verses). Maloy's edition of the offertory chants of the Roman and Gregorian traditions presents them side by side on facing pages on the book's website, hosted by Oxford University Press.Footnote ⁵ She made the justifiable decision to base the Gregorian edition on a consistent Italian tradition, that represented by the Beneventan manuscripts. Many Gregorian offertory chants have stable melodies with only minor variants, but significant variants (partial transposition, variable cadence tones, etc.) are present in no small number – a major challenge to attaining a comprehensive view of the genre. Chapter 6 (‘The Offertory in Manuscript’) presents a substantial essay on the manuscript basis of the editions. A ‘Selected Bibliography’ at the end of the book contains essential titles necessary to lead readers to fuller treatments of the various themes developed by the author.

Recent chant research has endeavoured to identify chronological layers within repertories. Since neither the Roman nor the Gregorian tradition provides obvious internal benchmarks for establishing chronology, attention has been directed towards the chant texts, more specifically to places where there appears to be a departure from the reading of the Roman Psalter (PsR), acknowledged to be the textual basis of the Mass chants.Footnote ⁶ It has been suggested that a number of chant texts (and presumably the original, no longer recoverable, melodies) originated outside Rome or display the influence of a non-Roman text tradition. Maloy favours these conclusions, which bear directly upon both ‘chronology and transmission’ (the book's subtitle). Of what help can these textual variants be?

The study of textual variants in the psalter and other books of the Bible is difficult enough for biblical scholars. Trying to compare biblical readings, on one hand, with chant texts, on the other, presents additional challenges.Footnote ⁷ First of all, there is the ‘apples and oranges’ problem of comparing two entirely different kinds of books, one scriptural, the other liturgical, that were created for completely different purposes and follow different principles of transmission. Medieval biblical texts were subject to emendation in order to achieve more accurate readings, limited necessarily by lexical equivalencies between Greek and Latin. Chant texts, on the other hand, were freely adapted to create suitable ‘lyrics’ (Maloy's felicitous alternative to ‘libretto’) for musical setting, thus often yielding what Philippe Bernard has called ‘liturgical variants’ that never existed in the psalter tradition. Insertion of the nomina sacra (‘domine’, deus') was common, as were omissions and rearrangements of verses.Footnote ⁸ Comparisons between psalters and chant texts are likewise hindered by the limited amounts of text in several of the six earliest (unnotated) chant manuscripts, published as Antiphonale Missarum Sextuplex by R.-J. Hesbert (Brussels, 1935). The Rheinau Cantatorium contains only the soloist's chants (gradual, alleluia, tract); the Senlis antiphoner provides only incipits. The other manuscripts contain longer or shorter amounts of text. Only the Antiphoner of Compiègne has complete texts throughout, fortunately including the offertory verses. Maloy is justified in devoting a long chapter (nearly sixty pages in length) to this subject, during which she grapples with several of the methodological obstacles.

Pre-Jerome Latin translations of the psalter are part of the ‘Old Latin’ (Vetus Latina) textual tradition, one strain of which coalesced into what is known as the Roman Psalter tradition.Footnote ⁹ After Jerome left Rome for Bethlehem in 385, he gained access to Origen's Hexapla edition, which allowed him to compare an array of Greek translations from the Hebrew. His revision of the Latin psalter based on these translations later became known as the ‘Gallicanum’ to distinguish it from the ‘Romanum’ in duplex psalters that presented both texts in parallel columns. Jerome's direct translation of the psalms from the Hebrew (‘iuxta Hebraeos’) diverges markedly in its vocabulary from the earlier Latin traditions; it never entered liturgical use. Only five Latin psalters can be dated before 600.Footnote ¹⁰ No psalter (or any other biblical manuscript, for that matter) has survived from Rome or central Italy that is contemporary with the creation of the Roman Mass Propers. Merely identifying individual words absent from Dom Weber's critical edition of the PsR is of limited value. For example, the offertory Super flumina is associated with α because in verse 1 the PsR reading ‘quia’ has been replaced by α's reading ‘quoniam’. Left unexplained, on the other hand, is why the very next word, ‘illic’, shifts back to the PsR and abandons the reading of α (‘ibi’).

The PsR tradition is not completely uniform, as the variant readings in the footnotes of Dom Robert Weber's critical edition, Le psautier romain et les autres anciens psautiers latins (Vatican City, 1953), demonstrate. Weber's edition has the advantage of comparing the PsR text with readings from ‘les autres anciens psautiers latins’, i.e., the wider Old Latin tradition, of which the PsR is a part. Such a parallel presentation is feasible because of the ‘parenté fondamentale’ between the Old Latin psalter tradition and the PsR. According to Pierre-Maurice Bogaert, ‘as different as they are, these [Old Latin] psalters can be collated with the Roman [Psalter]’.Footnote ¹¹ This factor needs to be kept in mind apropos of the text comparisons made in chapter 2 of Inside the Offertory. One cannot be certain, for example, that a reading in a chant text that departs from the PsR, but which is found in an Old Latin psalter from Gaul or North Italy, might not also have been present in a (no longer extant) PsR manuscript available in sixth-century Rome.

Maloy frequently references Andreas Pfisterer's claim that ‘variant’ readings came to Rome from the region of the surviving source. This claim is usually made with regard to a sixth- or seventh-century psalter in the Biblioteca capitolare at Verona (α in Dom Weber's edition), whose text form is identified as North African by Pfisterer, a view to which Maloy often subscribes, albeit with misgivings. The Verona Psalter was classified as an African text nearly a century ago by Paul (later Bernard, OSB) Capelle in Le texte du psautier latin en Afrique.Footnote ¹² He concluded (in italics for emphasis!) that ‘from a good thousand characteristic [readings] throughout the psalter, there are about 700 common to A[ugustine] and [Verona] and of these about 500 are found nowhere else’.Footnote ¹³ Even though this establishes the relationship between α and the psalm text used by Augustine in his Enarrationes in psalmos and other writings, it does not establish α as an African text tradition.

In a contribution to Alban Dold's comprehensive study of the sixth-century palimpsest psalter of St Gall (β in Weber's edition) Arthur Allgeier compared the text of Augustine's commentary on Psalm 118 with the text of the same psalm in α, thereby revealing quite a number of discordances between the two.Footnote ¹⁴ Allgeier also collated the text form used by Augustine's mentor, Ambrose, in his Expositio de psalmo cxviii. Of the 275 times in which the text of PsR and α differ in this psalm, Ambrose has 100 times the reading of α, and 125 times the reading of the PsR.Footnote ¹⁵ (Capelle had already identified fifty cases in which Ambrose agrees with Augustine only.) The conclusion must be that Ambrose, who had no strong African ties, cited a text readily available to him (and to Augustine) in North Italy.

Alberto Vaccari reviewed the researches of Capelle and Allgeier, comparing α with Augustine and with the psalter of St Gall (β), which was written at Bobbio, about 100 km distant from Milan.Footnote ¹⁶ Frequently α and β agree on the same reading (as can be seen from Weber's edition of the PsR and appendix 1 of Inside the Offertory), an agreement that reinforces the argument that α is part of a North Italian text tradition that may have once been more widespread on the Italian peninsula.

Even if the distant (3rd c.?) origins of the psalter type represented by α were African, it had become thoroughly ‘europeanised’ by the time Augustine would have encountered it at Milan. When he left North Italy to return to his native Thagaste (Numidia) in 388, the year after his baptism, he evidently took with him a North Italian psalter of the α type. Vaccari concludes that ‘it is necessary to clear from the table [rayer de la table] the words characteristic of Verona [and] Augustine, which Capelle qualified as “African”’.Footnote ¹⁷ Dom Bogaert agreed that the text type represented by α and β is ‘certainement originaire de l’Italie du Nord' and ‘n’était africain que par l'usage qu'Augustin et d'autres à son suite en ont fait de façon habituelle'.Footnote ¹⁸ Neither Augustine nor the Italian psalters of Verona (α) or St Gall (β) establish a ‘foreign’ influence on the chants of the Mass during their gestational period, and it is a not unreasonable assumption that a psalter of similar type was in use at Rome. Recourse to Africa is unnecessary.

Several times, moreover, α agrees with the sixth-century ‘Psalter of St Germain’ (Paris, Bibliothèque nationale de France, lat. 11947; γ in Weber's edition), a representative of the ‘gaulois’ psalter tradition. Very frequently, γ is joined by other psalters, including representatives of the PsR tradition.Footnote ¹⁹ Maloy properly discounts the significance of the ‘gaulois’ tradition for the offertories, doubting that ‘the surviving sources of the PsR and the gaulois traditions give us a picture of the early medieval texts that is sufficiently complete to serve as the foundation for a theory of origin’.

Not surprisingly, none of the presumed textual ‘variants’ has any ascertainable influence on the melodies: ‘the offertories with “foreign” verbal traits are musically indistinguishable from the repertory as a whole’ (p. 187). Acknowledging that ‘the conclusions to be drawn from these variants is, however, unclear’ (p. 37), she duly cossets her discussions with a multitude of caveats. She acknowledges that the variants between psalters and offertories are relatively few and asks ‘to what extent are the “non-Roman” verbal characteristics demonstrated by Pfisterer in a handful of offertories, found among the offertory repertory as a whole?’.Footnote ²⁰ The question becomes moot as soon as one recognises that the ‘non-Roman’ readings cannot be geographically localised and may very well be Roman after all. Maloy is far less cautious in the conclusions presented in ‘Origin and Chronology’ (ch. 5). There she claims that there has been ‘demonstrated a consistent “non-Roman” textual tradition in twenty-five of the Roman-Gregorian psalmic offertories and possible traces of non-PsR influence in at least two dozen others’ (p. 186). This does not seem to be the case.

Non-Roman origin has been claimed for non-psalmic offertory texts by Kenneth Levy, who based his conclusions about Gallican and/or Hispanic origin primarily on (in Maloy's words) ‘verbal variants and similarities’ (p. 61). She engages these theories numerous times, very helpfully clarifying for the reader the assumptions involved and her frequent dissent from them (pp. 57–67). After considering all the evidence from Spanish and Milanese sources, she seems to lean towards Rome as the point of origin for most of the non-psalmic offertories, which from a melodic standpoint are stylistically consistent with offertories of certain Roman origin. (Comparisons with the Old Hispanic tradition are made in ch. 4.) Some pieces may well be ‘of pre-Carolingian origin’, but direct proof is difficult to come by, since it is not known whether the Gallican sonum, sung as bread and wine were brought in solemn procession to the altar, was a single, invariable chant or a repertory that varied according to the seasons and feasts of the liturgical year. Maloy notes that chants, psalmic or non-psalmic, found in Gregorian sources but not in the Roman manuscripts may merely have fallen out of use in the Eternal City between the eighth and the eleventh century.

Analysis of the music of the offertories (ch. 3) begins with the Roman corpus and its most obvious feature: the presence of reciting formulae in responds and verses.Footnote ²¹ Maloy concludes that ‘the Roman formulas are consistent with the model of oral transmission proposed by Leo Treitler’, and her insight that ‘[the singers] absorbed the principles of making formulaic melodies intuitively and nonverbally in the process of rehearsing individual instances of the formulas’ (p. 96) is certainly correct. She recognises that a stylistic continuity exists between these formulaic Roman melodies and the ‘idiomelic’ ones. While long passages had to be learnt by rote memorisation, other portions were formed by a ‘reconstructive process’, the difference due largely to ‘the degree of stereotyping present in their surface details’ (p. 103). While Maloy expresses admiration for the Roman singers' skill in matching the formulae to syntax and word accents,Footnote ²² the descent of the melodies into ‘formulaicism’ will be seen as a failure to preserve the eighth-century tradition. One wonders how the (ostensibly banal) Roman formulae were really performed: simply ‘come stà’ (as Arcangelo Corelli specified in one of his violin sonatas) or – more likely – with a rhapsodic freedom that animated the melodic line.

Gregorian offertories are more expansive in range and more individual in their melodic contours and placement of melismas, some truly striking in their flamboyance. Repetition plays a much smaller a role, apart from the text repetitions that are so conspicuous a feature of the offertories of both traditions. Maloy exemplifies some of these aspects in a comparison of the Gregorian and Roman versions of Iubilate deo universa and Domine fac mecum and places them in a theoretical framework with her discussion about ‘defining similarity and difference’ (pp. 106–10). This leads to a mapping of both repertories according to the degree of resemblance between the two, ranging from 1 to 5 (Table 3.1), the weakest relationship (5) registered in those Roman offertories dependent on one of the formulae. Advent and Christmastide verses exhibit a higher degree of affinity (category 2 or 3) than do half the Lenten verses and most of the Paschaltide ones (category 5). Maloy regards this tendency as an indication that ‘after the separation of the two traditions, the offertory verses were subject to increasing melodic change in the course of the liturgical year’ (p. 121). Why this should be so is not entirely clear. A roughly similar pattern is found in the prevalence of non-formulaic pieces during Advent and Christmastide in the Roman tradition compared to the more extensive appearance of formulae from Epiphany onwards. How this is to be explained is left to a later chapter, but a brief note on Domine fac mecum, which employs the ‘torculus’ formula A, makes the point that it cannot be a pre-existent Gallican melody, as hypothesised by Kenneth Levy. Presumably the same conclusion applies to other texts set to the same formula.

Chapter 4 confronts problems presented by ‘The Milanese Melodic Dialect and the Question of “Italianate” Style’, an important topic inasmuch as nearly half the Milanese offerendae are closely related to the Gregorian repertory.Footnote ²³ Their musical adaptation appears to fit comfortably into what is considered to be a typically ‘Italianate’ idiom. Examples 4.1–3 illustrate well the melodic relationships among the three chant families, but the detailed overview in Table 4.1 indicates that Milanese liturgical assignments are relatively independent.Footnote ²⁴

Maloy asks where the Milanese offerendae with Roman-Gregorian cognates originated and when they were incorporated into the Milanese repertory. She is not inclined to accept the consensus that they were borrowed from the Frankish Gregorian repertory at a relatively late date and then adapted to the local musical style, but neither can it be the case that a ‘similar melodic vocabulary emerged independently in Rome and Milan’. Maloy compares the ‘ornamentation’ of offerendae and offertories, but the evidence seems to point in several different directions with respect to their relationships. The Milanese chants, as the Gregorian offertories, know nothing of the formulae of Roman chant. A handful of cases suggest some acquaintance with aspects of the Roman musical tradition, such as the (Roman) melisma on the first syllable of ‘induit’ in Deus regnavit (Example 4.2) and the absence of the (Gregorian) melisma on the last syllable of the previous word (‘decorem’). She attaches decisive importance to divergences that agree with Roman melodic readings.

A comparison of the Milanese Oravit Moyses with the Gregorian and Roman melodies and a sacrificium from the Visigothic Antiphoner of León (unheighted neumes) reveals similarities and differences (Example 4.5). The Old Hispanic tradition lacks the long melisma on ‘Oravit’ of the Gregorian and Milanese traditions, and the amount of surface detail differs, but Maloy affirms ‘a close relationship between the Old Hispanic and Milanese’ vis-à-vis the Gregorian and Roman cognates. If these and other offertories had been transmitted in a formulaic state from eighth-century Rome to Francia, Maloy finds it difficult to explain ‘the greater melodic correspondence [of the Gregorian melody] to the Milanese and Old Hispanic readings’. She proposes instead that ‘the eighth-century Roman tradition had melodies for these offertories that were similar to the Gregorian versions but much of the Roman melodic tradition was lost between the eighth and eleventh centuries’ (p. 177).

Chapter 4 closes with a brief reflection on Roman ‘formulaicism’, in which Maloy rejects the notion that the Old Beneventan and Roman offertories exhibit a Italianate style that reaches back to the sixth or seventh century, preferring instead the hypothesis that during centuries of oral transmission they developed a similar ‘melodic economy’. The process is described as a ‘tendency for idiomelic melodies, especially verses, to become increasingly repetitive and formulaic in the Roman tradition’ (ibid.). She modifies this position somewhat with the admission that ‘it would be imprudent to attribute all Roman and Beneventan characteristics, including melodic thrift, to later development’ (p. 181).

Chapter 5 (‘Origin and Chronology’) summarises findings of previous chapters and reiterates their conclusions. Maloy endorses an ‘early’ scenario against James McKinnon's hypothesis that the chant repertory was created or substantially revised by members of the schola cantorum in the later seventh century. She would place the terminus ad quem much earlier in that same century. A brief review of offertories whose origin may be ‘non-Roman’ (a term very frequently enclosed within quotation marks) remains to an extent inconclusive, since even pieces that might be thought to have ‘non-Roman’ or ‘pre-Carolingian’ origins (Stetit angelus, Elegerunt, Factus est repente, Audi Israhel, Ingressus est Zacharias) are ‘stylistically indistinguishable from the standard Gregorian offertories’ (p. 189). Nevertheless, this comment initiates an illuminating discussion of the latter two pieces, which leads to the conclusion that the chant found in the Roman manuscripts is isolated from the ‘international tradition, incorporating certain stylistic features common to Spain, Gaul, and northern Italy’ (p. 192).

The author thus proposes that the offertories transmitted from Rome to Francia towards the end of the eighth century and preserved in East and West Frankish manuscripts of the ninth century and later represent not a Frankish reworking of an imported repertory but rather Roman chant of the period: the ‘Gregorian’ melodies are substantially what was being sung in Rome at the time of Charlemagne (and hence cannot be called ‘Romano-Frankish’).Footnote ²⁵ This view rejects the current scholarly consensus that ‘Gregorian’ chant is a Frankish redaction of a received Roman repertory. The offertories found in the Roman manuscripts (and a fortiori the entire repertory of which they are a part) are the result of centuries of oral transmission during which most of the melodic details of the seventh- and eighth-century offertory melodies and their verses were largely forgotten, replaced by a ‘progressive homogenization’ that produced only a ‘perfunctory formulaicism’ (p. 143). In this view, the Roman singers cut a very poor figure: melodies that in the eighth century were highly individualised collapsed over the course of time into the stepwise, weakly articulated melodies of the eleventh- and twelfth-century manuscripts.Footnote ²⁶ Far from representing an ideal means for remembering melodies in an oral environment, the two offertory formulae and the (apparently artful) repetition techniques represent a failure to preserve the melodies. The large tracts of the Roman offertories composed of ‘original melodies’ are considered to be passages in which the authentic melodies, preserved in Carolingian manuscripts, were disfigured, but not yet reduced to the level of fomulaicism. Apparently, Milanese singers succeeded rather better in preserving the melodies of that portion of the eighth-century Roman offertory repertory that they accepted.

The idea that Roman singers corrupted the chant is not new. It was Dom André Mocquereau's first reaction to the discovery of a parallel body of music, genuinely Roman, that differed markedly from the repertory he was striving to recover and promote as the unique musical heritage of the Roman Catholic Church, Gregorian chant. Bruno Stäblein, on the other hand, asserted sixty years ago that this ‘altrömischer Choral’, as he called it, could not be so easily dismissed, thus initiating a new era in chant research. Multiple theories were developed on both sides of the Atlantic to explain the situation. These necessarily reflected on mechanisms of oral transmission, and chant scholars gained a respect for the tenacity of memory – important as well for unheighted neumes – a respect that Maloy inclines to believe might have been too hastily and uncritically awarded.

Maloy's conclusions imply scepticism about reports (with which she is familiar) that the Franks had difficulty learning the chant taught to them by the Romans, and that it was sung differently in various parts of the Carolingian realm. Bishop Rémy of Rouen (755–72), brother of King Pepin, had to send some of his clerics to Rome for instruction. In adjudicating a dispute between Frankish and Roman cantors, Charlemagne also decided that the music could best be learnt atRome (‘revertimini ad fontem’). It could be countered, of course, that, despite considerable challenges, Frankish cantors did accomplish what Charlemagne commissioned them to do in one of his capitularies – ‘ut cantum romanum pleniter discant’.

The idea that Frankish singers accepted without alteration ‘their’ component of the Roman liturgy needs to be viewed against the larger (and different) picture of the ‘échanges liturgiques’ (Cyrille Vogel) that took place between Rome and Francia in the late eighth century.Footnote ²⁷ The received liturgical books had to be adapted to situations vastly different from those prevailing at Rome. Though the Creed and the baptismal rite remained sacrosanct, nearly everything else was open to modification. Even the Gregorian Sacramentary, believed to be the legacy of an admired pope (‘a sancto Gregorio papa romano editum’, according to its title), could not lay claim to iconic status. Almost immediately upon receipt it was supplemented with Masses and prefaces from the Frankish tradition. Neither was the Mass lectionary immune from adaptation, as Theodor Klauser and Walter Howard Frere have shown.Footnote ²⁸ Amalar of Metz, though respectful of Roman practice, assumed that he could produce an Office antiphoner on his own authority to reconcile the inconsistencies between Roman and Frankish practice.

If the melodies were learnt with exactitude by Frankish cantors, it is puzzling why the original PsR texts were emended in dozens of places to bring them in line with the hexaplaric (Gallican) psalter, then coming into favour in the Carolingian realm.Footnote ²⁹ The idea of an integral acceptance of the Roman melodies may also need to address the disparate amounts of text present in the earliest graduals published by Dom Hesbert. Finally, there must have been a measure of hostility to the Carolingian programme of Romanising. At the conclusion of Ordo Romanus 19 its compiler expresses forcefully his indignation at those who would reject Roman practice by claiming that they followed the traditions of Hilary, Martin, Germanus or Ambrose, esteemed bishops who would never ‘deviate from the Holy Roman See or from the 318 Catholic Fathers of the Council of Nicaea’.Footnote ³⁰

Inside the Offertory is a very complex book and a review (even a rather long one) cannot do justice to the extraordinary amount of material or to the thoughtful reflections thereupon. Though one may have reservations about conclusions that can be drawn from psalmic texts and comparisons with repertories that cannot be transcribed (Old Hispanic) or have substantially disappeared (Gallican), Maloy draws on this material judiciously and with full command of previous research. High on the list of lasting contributions of this monograph will surely be its side-by-side edition of the Roman and Gregorian repertories on the Oxford University Press website, supported by the detailed commentary on every single piece. May this major contribution to the study of the offertory inspire similarly detailed books on the other Proper chants of the Mass.Footnote ³¹