This is a book that inspires confidence from the outset, with a well-considered summary of the historical background to the transmission of the chant from Rome to Francia and of the liturgical history of the offertory in Rome and elsewhere. It continues with chapters on the offertory texts, the musical characteristics of the Gregorian and Roman offertories, and those of the Milanese melodic dialect, followed by a résumé of the arguments for the relationship between the Gregorian and the Roman repertories and a final chapter on the manuscript evidence by way of introduction to the edition. The appendices provide in turn a summary of the textual sources and variants, a list of manuscripts cited, and finally a commentary on the edition of ninety-four offertories that can be found as pdf files on the associated website.

This is a complex and thoroughly researched book, but its main contention can be stated quite simply. It is that the parts of the Roman melodies that rely on formulae, and which differ most from their Gregorian counterparts, represent a development in Rome that post-dates the transmission of the melodies to Francia. This will seem counter-intuitive to some, and it contradicts the assumption made in most of the recent literature on the subject, namely that formulaicism is an inherently archaic feature of chant, discarded in favour of more ‘artistic’ idioms as the inventive spirit of cantors and their scholae came to the fore. I have to say, however, that I am completely convinced by the general picture that emerges from this study, and at the risk of trivialising complex arguments I shall try to explain why.

Throughout the book, Maloy's assumption is that the surviving Roman and Gregorian repertories of chant (her terms) descend from a common Roman ancestor. I will not quarrel with this, since it is now generally accepted, but although she does adduce evidence in support, the hypothesis is not fully argued here and has not always been adopted. Originally – and here the wheel seems to be coming almost full circle – the Roman manuscripts were thought to represent a decadent form of Gregorian chant. It was subsequently argued that on the contrary they represented an older state, and that the Gregorian melodies were a revision of them; and because Gregorian chant was both truly Roman and truly Gregorian, the Roman versions must be pre-Gregorian, and in this sense ‘Old Roman’. Even when these chronological assumptions were discarded, the idea that the one (Gregorian) was a conscious revision of the other (Old Roman, surviving miraculously in sources no earlier than the eleventh century) continued to dominate discussion. The argument then turned on the circumstances which might have produced two similar but distinct chant dialects within the one city. Only later did it become clear that the Gregorian dialect, surviving to begin with almost exclusively in Frankish manuscripts, must be a Frankish repertory that in due course superseded the original Roman repertory in its own city. The idea that it is very much a Frankish creation, influenced by and incorporating earlier Gallo-Frankish chant idioms, has led many scholars to refer to it unapologetically as Frankish (in contradistinction to Roman) chant. But this conflicts with the historical evidence that the Franks at least intended to preserve Roman chant as an integral element of the Roman liturgy that in principle they also adopted; and I think Maloy is right to retain the less tendentious term ‘Gregorian’, despite its historical falsity. She does, however (like others), drop the ‘Old’ from Roman, and in view of her findings it is not hard to see why.

Maloy's principal argument in favour of a later development in Roman chant as we have it is very simply that it (apparently) relied on a purely oral tradition for a much greater length of time than the Gregorian, which first received comprehensive notational form around 900. She allies to this essential truth an interpretation of the evidence of the formulaic passages in the Roman offertories that is novel but persuasive. It is important to realise (though she does not emphasise this) that evidence of this kind, should it be accepted, enables one to apply its implications to parts of the repertory which do not exhibit it.

Briefly, Maloy argues that formulaicism, so far from representing an archaic practice, emerged as a substitute for the memorisation of elaborate melodies that had not been systematised to anything like the same extent. She demonstrates (and this is the main new contribution that she makes to the understanding of the Roman melodies) that the formulae of the offertories were applied in a detailed and rigorous fashion to the accentuation and syntactical structures of their texts. Others have demonstrated the existence of the two principal such formulae, and have indeed recognised aspects of their textual basis; Maloy has noted others, some of which are specific to a single mode, and has described in detail how the formulae were applied. She observes that the formulaic passages of the Roman offertories, which are concentrated in their verses (rather than the responds) and in the later parts of the liturgical year, are those which least resemble the corresponding sections of the Gregorian offertories; and she wonders why, if the Frankish cantors were concerned to adopt Roman chant, they should have ignored the passages allowing a reconstructive process of memorisation rather than the more challenging task of rote learning. Some musicologists, indeed, have argued that the adoption of Roman chant was not the aim of the Frankish cantors, but she finds that the historical evidence is against this view, and again it is hard not to agree.

Although the main purpose of this book is to support its thesis with the available evidence, and thereby to delineate an essentially new approach to the understanding of Roman–Frankish musical relations, it does much more than that. On the one hand, it reviews all the pertinent material relating to the liturgical history, textual basis, and melodic relationships between the offertories of these and other chant repertories; on the other, it provides a comprehensive edition of the Gregorian and Roman offertories in such a way that they can be directly compared, and so far as the Gregorian melodies are concerned with a wealth of text-critical sophistication on the basis of a multiplicity of sources.

There are of course larger questions concerning the origins of the conventions within which all parts of the Western chant repertories arose: questions of textual choice, style, modality and form in relation to the liturgical context of each genre. Many of these issues have been addressed in recent scholarship. So far as the offertory is concerned, Maloy's attention, like that of others before her, is focused initially on the early history of the rite as well as on its possible musical accompaniment. I find her observations refreshingly candid. For example, in regard to the ‘hymni … ante oblationem’ that were defended by St Augustine (according to an exchange referred to in his Retractiones) against the objections of a certain Hilarius, she is ‘hesitant’ to exclude the possibility that this expression referred to an offertory chant (though not, evidently, in the form that it later assumed). There is good reason for hesitation, since oblatio can mean many things, from a donation by a lay person to the formal presentation of bread and wine on the altar (the word is the equivalent of the Greek anaphora, which can also mean more than one thing but is most commonly associated with this last act and the remainder of the eucharistic action). Other interpretations have necessarily been explored, but the one that Maloy is hesitant to exclude seems in the end the most natural one.

Difficulties also occur in interpreting the Gallican offertory rite and its accompanying sonum, as described in the anonymous Expositio brevis.Footnote ¹ As she says, this was modelled on the Great Entrance of the Eastern rite, but it includes a puzzling reference to the oblation as corpus domini and sanguis Christi. This cannot be taken at face value, and Duchesne's brief comment seems quite adequate: ‘On lui rend par prolepse les mêmes honneurs qu’après la consécration.'Footnote ² More relevant, perhaps, is Maloy's conclusion that ‘little can be discerned [from the Expositio] about the nature of the Gallican offertory chant’. In the first of the Ordines romani the chant that accompanies the rite is indicated obliquely by reference to a signal from the pope that it should stop (‘annuit ut sileat’, though this is followed by ‘Tunc, finito offerturio’, presumably a reference to the chant rather than just the rite).

Discussion of the early history of the rite can take us only so far in understanding that of its accompanying chant. The same is true of the study of its texts, for the very diversity of their musical settings in Roman and Frankish sources, together with a number of Milanese cognates, argues for a prehistory that cannot be securely recovered. Nevertheless the texts can tell us much about the growth of the repertory and the way in which it was perceived. Previous scholars have emphasised the centonised nature of the chant texts, and more recently Andreas Pfisterer has conducted an exhaustive examination of the biblical translations used in the compilation of all the chants of the Mass.Footnote ³ Maloy is an excellent guide to all this, pointing out, however, that the supposed replacement of one biblical version by another, in Rome or elsewhere, is not an infallible guide to the dating of a text. Those who created the texts may have had conservative preferences, or relied on imperfect memories. Moreover, the oldest known Psalter versions are often represented by only a single source, or to put it another way have been named by textual critics on the basis of a single source. For the rest of the Bible, the pre-Vulgate versions known collectively as the Vetus Latina have had to be reconstructed from a variety of sources. Again it is hard to disagree with Maloy's remark (p. 77) that ‘the complexity of the Bible's textual history suggests that this evidence should be employed with some caution as a means of dating specific chants and the repertory as a whole.’

The possible origin of the offertory outside Rome has also been the subject of informed speculation. A connection with the Hispanic sacrificium has long been suspected; its texts are predominantly non-psalmic and are usually concerned with the sacrificial rites of the Hebrews as a ‘type’ of the eucharistic sacrifice about to be enacted. It has been natural to focus on the few non-psalmic offertories of the Roman rite in the search for links, though not all the non-psalmic offertories are ‘sacrificial’ and only one text actually corresponds to that of a Hispanic sacrificium. Kenneth Levy has proposed that the form originated in the Gallican sonum and spread from there to both Spain and Rome;Footnote ⁴ but we do not know enough about the texts of the sonum to be certain of that, and it remains an intriguing hypothesis.

Perhaps greater significance attaches to the formal similarity between the Hispanic and Roman genres, in which a principal section is followed by one or more verses, each (in principle) concluding with a portion of the main section generally referred to as the repetendum. One of the differences is that the Hispanic form regards the principal section as the first verse, numbering the remainder as ‘ii’ etc. (up to as many as five in one instance). In Roman and Frankish manuscripts the verses are not numbered, but the Ordines and other sources clearly distinguish between the principal section and the verses. In neither the Hispanic nor the Roman/Frankish forms is the indication of the repetendum always present; and the point from which the repetition is to be made is not necessarily the same after every verse. Nor is the identification of the start of each verse always identical between the Roman and the Frankish manuscripts, or even, to a lesser extent, within the Gregorian tradition.

The question has often been raised as to whether the offertory chant is antiphonal, responsorial, or neither. When Apel's view that the offertory had originally been provided with verses sung to simple psalm-tones such as the introit (and communion) was shown to have been based on a misunderstanding of a passage in Aurelianus Reomensis, the notion that it was ‘antiphonal’ was universally discarded; yet in a wider sense we could perhaps think of it as an antiphon with verses, and certainly as being analogous in function to those action-chants. James McKinnon insisted that the offertory was sui generis, and not a responsory,Footnote ⁵ while Maloy with justifiable pragmatism uses the vocabulary of respond with verse or verses; consequently her descriptions are entirely unambiguous. At the same time the music of the offertory, at least in the Gregorian versions, tends to exhibit a progression from the relatively straightforward to the more complex as the sections proceed rather than a clear contrast between the respond on the one hand and its verses on the other.

There has also been a general assumption, arising from the ‘responsorial’ interpretation and shared by Maloy, that the principal sections of offertories were sung chorally and the verses soloistically. This too, I suggest, is a historiographical convention rather than something that can be demonstrated beyond doubt. There are passages of verbal repetition that make one wonder whether some kind of alternation of forces within a section might have been intended. In any case, the musical manuscripts contain no rubrics. Moreover, some texts give the impression of a more or less continuously evolving narrative rather than a series of formally discrete sections, in which there is no obvious verbal reason for a repetendum, though it would normally be required for the sake of tonal coherence.Footnote ⁶ A good example would be ‘Vir erat in terra Us’, in which the verses, giving the words of Job in direct speech, follow an introductory narrative. This is also a case where the verbal repetitions, especially the ninefold ‘ut videam/videat bona’ in the final verse, could conceivably have been sung by cantors in alternation, creating a quasi-operatic echo effect. In the Vatican manuscript of Roman chant, each of these repetitions is shown as a separate verse.

Here, in most versions of the chant, a repetition of the respond is needed, at least after the final verse, if the tonality is to be restored. The cue in Einsiedeln 121 and other manuscripts is to the beginning of the respond, ‘Vir erat’ – but perhaps it is not necessary then to proceed to the very end: a repetition of the first phrase would suffice for that purpose. But once the melodies are taken into consideration, further questions arise. We know that originally, at any rate, the pontiff gave a signal to the schola to stop singing once the offerings had been collected (so the chant was not originally intended to cover the oblatio at the altar). The singing might not have come to a halt immediately, but on the other hand it seems likely that formal or tonal coherence was regarded as expendable (plus ça change …). Further, the melodies as preserved show more than the usual indications of tonal instability – variant pitch levels for entire chants, for sections of chants, and for shorter passages within sections. The reasons for these differences of transmission are perhaps the most interesting aspects of this challenging repertory.

Maloy's analysis of the melodic characteristics of these chants, and the comparisons she makes both between the Roman and Frankish versions and between these and their Milanese cognates (where these exist) and even, in a few cases, with uncleffed Beneventan and unheighted Hispanic material, form the core of the book and are her most impressive achievement. In her editing of the corpus she presents on alternate pages the Roman and Gregorian versions and a full commentary on the variants within each. It is the Gregorian versions, with their enormously larger number of manuscript witnesses, that call for by far the fuller treatment, and it is in this treatment that the issue of written pitch-levels comes most sharply into focus.

Other recent writers, notably Karp and Atkinson, have drawn attention to the use of transposition to accommodate pitches not available in the standard Guidonian scale that forms the basis of the cleffed staff. The relocation of whole chants to finals on a, b or c is familiar enough and discrepancies between offertory responds and their verses have long been evident; what becomes clearer from this analysis is the extent to which pitch-instability pervades the entire repertory, and the difficulty of arriving at authoritative versions when one is faced with a multiplicity of choices. Maloy's editorial methods are pragmatic: she chooses a base manuscript, usually a Beneventan or an Aquitanian source, that in her view represents the most probable version, though with editorial accidentals at points where the differing pitch-levels of other manuscripts suggest that they are needed. Some flexibility is offered by the use of diastematic but uncleffed sources, and the choice of base-manuscript is sometimes suggested by the pitch-level of the Roman versions. However, she is content to present the evidence rather than to offer more radical (and hypothetical) relocations of sections of chant. The evidence is suggestive rather than conclusive, and she is often willing to say that she does not know the reason for a particular variant.

The whole exercise deserves a more detailed and sophisticated response than it would be practicable to offer in this review. At the deepest level it invites an imaginative reconstruction of the processes of chant composition in what has been termed the pre-theoretical age. Unfortunately it is impossible to say exactly when chant theory came to determine its melodic form. We know that an eight-mode system was known in Francia from around 800; I have argued that this is a logical outcome of the D–d scale, arranged as a pair of disjunct tetrachords, common to Greek and Latin chant theory, and Michel Huglo has demonstrated the knowledge of that octave-scale in Hispanic sources of Isidore's Etymologiae at around the same time. So it looks as though this would have provided a framework for musical expression, possibly for some time before that, and perhaps also a rationale for procedures that were largely instinctive in any case. But it need not have created the restraints later imposed by the Guidonian scale, and the evidence for the use of intervallic structures that the latter could not accommodate is too strong to be ignored. Maloy uses the term ‘non-diatonic’ to describe these; while this is an improvement on the now discredited ‘chromatic’, it has a similar implication, and I would prefer the term ‘non-systematic’. Except for the possible use of microtonal intervals, the procedures involved could perhaps be thought of as modulations arising from a shifting of tetrachordal positions: that, at any rate, is what the evidence of temporary pitch-variants as well as the later theory of coniuncturae would suggest.

Another focus of speculation, arising from but not restricted to a consideration of the offertories, concerns the process by which moderately and highly melismatic melodies came to dominate the Mass repertory. Assuming that the liturgical forms were originally much simpler in style, and at least partly congregational, the development must have gone hand-in-hand with a radical change in the nature of Christian worship, from a fundamentally participatory conception to one in which the verbal content of the liturgy was proclaimed by specialist performers on behalf of the congregation. This would have affected especially the psalmody of the fore-Mass, but it must have influenced the development of the ‘action-chants’ as well (chants which because of lay involvement in the action or later origin, or both, were the special province of dedicated singers).

The comparisons undertaken by Maloy across the repertories, and especially when the Milanese cognates are taken into account, make it clear that comparable but differing processes of elaboration have left their mark on the written sources. These comparisons need not be limited to items of equivalent liturgical function, and I would have been interested to have Maloy's observations on the Roman/Gregorian offertory ‘Ad te domine levavi’ and the clearly cognate Milanese ingressa and confractorium for the first of its six advent Sundays. The confractorium shares the text of the verse ‘Dirige me’ of the Roman offertory, and moreover concludes with a repetition, from the word ‘etenim’, of the last part of the ingressa. One could say that these Milanese items were ‘borrowed’ from the Roman repertory and applied to these different liturgical functions; but if so it will have been at a time long before the melodies had reached their final shape in any of these chant dialects, and also perhaps at a time when Roman melodic style was less specific to a particular genre than it later became.

It is tempting to create hypothetical simple originals from which this and other groups of related melodies originated, but the process of elaboration must have been more complex than the mere melismatisation of a basic framework. Different chant repertories, and different genres within them, will have employed different processes of extension as well as having different ends in view. In that context one can imagine that the differences between Roman and Frankish cognate chants do not reflect wilful or even accidental departures from a purer original, but are the consequence of distinct and equally legitimate ways of handling the raw material. To put it another way, when Frankish cantors adopted Roman chant, it was probably never expected that this would have amounted to a note-for-note reproduction: the tradition of oral composition was still firmly entrenched, and the relative standardisation of the written Frankish repertory will have been due to a widely shared understanding within the Frankish Church as to how the Roman model might be creatively interpreted. Later references to the supposed incompetence of the Frankish singers could well be attempts to account for differences from the standpoint of literal rather than exemplary transmission.

In the context of the offertories this allows us to understand better the use of strictly controlled formulae in the Roman versions, particularly in the final verses and in the later parts of the liturgical year. With the passage of time, and during a longer period of oral transmission, a more systematic method of elaboration may have seemed preferable for one reason or another, and given more time might have been more widely applied. But the fact that it affected only parts of the repertory allows us to see that it was not an original feature of the Roman aesthetic; the alternative would be to believe that the Roman cantors embarked on a project of idiomelic recomposition which made their offertories more like, but not exactly like, the independently developed Gregorian melodies. Gregorian offertories also share related material amongst themselves, but not in the same systematic fashion.

I have only a few criticisms of this book, and most of them are as much to do with the publishing process as with the authorship. There are a very few misprints and bibliographical oversights, including an amusing reference to the Larry Bradshaw Society. On one substantive point, I would like to have seen Maloy's comments on Joseph Dyer's recognition of the Roman offertory formula A in the (Latin) alleluias for Vespers during the Easter octave.Footnote ⁷ Otherwise it is largely a question of helpfulness to the reader. There is no actual list of the offertories with Maloy's numbers. The commentary on the edition acts as a kind of list, but it is spread over nearly 180 pages and does not indicate the liturgical assignments. The table of non-psalmic offertories on page 58 gives the assignments but not the numbers. Appendix 1 gives the numbers of the psalmic offertories and (most usefully) the textual variants from a range of psalters, but it omits offertories in which there are no variants to record. The website lists and numbers the offertories concisely as a set of pdf files, but does not include liturgical assignments, either in the list or within the files themselves.

It would certainly have been useful to include such a list in the book itself, perhaps with cross-references to the numbers in Rupert Fischer's Offertoriale triplex (1985), itself based on Karl Ott's Offertoriale of 1935. Inadequate though this publication undoubtedly is, it is a resource of a kind, and many of us will have it on our shelves and perhaps even have sung from it. I have referred to it frequently, as well as to the more reputable editions of Montpellier H 159 and Vatican lat. 5139, while consulting Maloy's commentaries.

A further quibble is that while there is a list of manuscripts used for the edition, there are some other abbreviations used earlier on – for example, for the manuscripts of Hesbert's Sextuplex – which for perfectly good reasons are not those of Hesbert himself. These are explained on their first appearance, but not subsequently, which can be inconvenient; a more comprehensive list at the beginning of the book would have been more helpful. As for the commentary itself, it would again have been useful to have had all the manuscripts used for a particular offertory listed at the head of each item. Those that involve variant readings are discoverable during the course of the commentary, as is (in most but not all cases) the base manuscript selected. The latter is nevertheless always indicated at the head of each transcription.

My main complaint about this publication, however, is one for which the author cannot be held responsible and with which she may well agree, namely that her edition is confined to a series of pdf files on the associated website. This is free to the user, who can access it simply by means of a user-name and password printed on the title verso of the book. This free access might well be considered a benefit outweighing any disadvantages, but I for one have found it impossible to browse efficiently across the repertory. (The website does not have any interactive functions.) For real comfort one would have to print it all out, an immensely time-consuming task requiring large quantities of paper, since the layout is generous, with no more than eight staves to the page. A printed companion volume in normal folio format could have been far more economical with space, and not prohibitively expensive.

As for the edition per se, it is an immensely valuable contribution to chant scholarship, particularly for its comprehensive treatment of the Gregorian part of the repertory. The commentary, while not quite exhaustive, is as detailed as anyone is likely to require, and is frequently illustrated by lengthy musical examples setting out variant readings. The Roman side of the equation needed much less commentary, given that there are only three manuscripts, of which the Bodmer Codex lacks a good part of the liturgical year and the later San Pietro F 22 lacks the verses of the offertories.Footnote ⁸ Even so, all significant variants are recorded. In the last resort I should have preferred a modified square-note transcription on the four-line staff – since the comparisons are lateral rather than vertical the necessity for note-for-note equivalence is less evident – but the stemless note heads employed here are nevertheless clear and unambiguous.

This publication is sure to find a warm welcome amongst chant scholars everywhere. It immeasurably advances our knowledge of the Roman and Gregorian offertory and thereby greatly enlarges the perspective within which Roman/Gregorian relations might be considered.