The earliest example of musical notation described in this Catalogue was written at Reims in the 880s [H34]:Footnote ¹ the notation, for one line of text only, is in ‘Messine’ neumes.Footnote ² Comparison with other sources copied in the area of north-eastern France bounded by Laon, Compiègne, Soissons and Reims confirms this as the standard notational type written in this region in the late ninth and early tenth centuries.Footnote ³ Musical notation may already have been in use for a substantial period in this region, possibly since the middle of the ninth century: that at least is a possible way of reading the evidence of not only a good number of late ninth-century examples, but also the highly developed nature of some of these notations, above all in Laon, Bibliothèque municipale 239.Footnote ⁴ One of the sources used to provide neumatic notation for the Solesmes Graduale Triplex,Footnote ⁵ the Laon Gradual contains some of the most detailed and subtle notation ever made for Gregorian chants. That book may have been copied as early as the 880s, and certainly before 900.

If the extant evidence is to be relied on, the period in which musical notation began to be used in England was significantly later. Although this catalogue includes other examples of notations possibly written in the ninth century, none can be securely demonstrated to have been written in England; equally, examples of notation associated with material copied in England, or with books owned in England in the ninth century, cannot be shown to have been entered before a later period.Footnote ⁶ Possibly the earliest relatively substantial example of notation written in England is on two folios from a gradual used as endleaves in Cambridge, Pembroke College 46 [H48]: here, texts for a series of Alleluias with verses were laid out with spaces to allow for melismatic passages of notation. A first layer of notations on these pages is directly associated with the text hand, which can itself be dated roughly to the first half of the tenth century.Footnote ⁷ A second layer of notations, some added where chants were not notated by the first notator, and some over erasures of the first layer, are of an entirely different kind. The first set of notations uses the so-called ‘Breton’ neumatic notation,Footnote ⁸ dominated by individual points for individual pitches, while the second uses ‘Anglo-Saxon’ neumatic notation, dominated by strokes which may indicate multiple pitches. That the one type replaced the other is evident; whether that choice was made through change of practice in one centre, or through movement of the host book from one centre to another, cannot now be known. Nevertheless, the pages are of foremost interest, since the pattern of use and change they reflect can be documented more substantially through other sources of the later tenth and early eleventh centuries: moreover, their value does not depend on identification of an insular origin, since, if the first layer of notations was copied on the Continent, we can see precisely what an Anglo-Saxon musician would have seen – whether or not he had learnt of such notation before encountering this example.

The four pages of Pembroke 46 summarize the history of musical notation in England between 900 and the early eleventh century. Notations of the Breton type can be found in fifteen books made in England or brought to England in the tenth or early eleventh centuries.Footnote ⁹ Of these the most extensively notated examples include the Pontifical Paris, Bibliothèque nationale de France lat. 943 [H310] and a collection of saints' lives and offices in London, British Library Harley 1117 [H157]; in the former the notations were not made at the time of the book's copying (in square minuscule) but later, at the time of its revision (in English Caroline), probably in the first decade of the eleventh century. Notations in Harley 1117 also date from c. 1000. The secure association of Harley 1117 with Christ Church Canterbury (through an identified text scribe) is the central argument for a link between this type of musical notation and Canterbury.Footnote ¹⁰ The use of Breton notation with marginal annotations (sets of chant incipits) made to the Leofric missal while it was in Canterbury (920–1000) underlines the link.Footnote ¹¹ Yet by the second quarter of the eleventh century Canterbury had relinquished its use of this type of notation and gone over entirely to the ‘Anglo-Saxon’ type, already practised at Winchester.

While there are four Anglo-Saxon examples in which Breton notation is superseded by Anglo-Saxon,Footnote ¹² there are none of the reverse: it was to the newer way of writing notation, developed in England but closely based on north French models (especially from Corbie) that Anglo-Saxon musicians turned. While the implication of the earlier Breton examples in Anglo-Saxon books is that musical notation was present in Anglo-Saxon England before the middle of the tenth century, the introduction of models from Corbie, and the development of the Anglo-Saxon neumatic script, was inspired by the Benedictine reform which gathered momentum in the 960s; of especial interest is the introduction by Æthelwold of monks from Corbie to Abingdon in the period between 955 and 963.Footnote ¹³ The pattern of use and development of the two musical notations has clear parallels to the development of new text scripts in late Anglo-Saxon England: by the late tenth century the two Canterbury houses had developed related new kinds of Caroline script, possibly based on Continental models, and characterised by its small size and elegance, while at Winchester another new, but different, kind of Caroline script, ‘broad and rounded, with excellent proportions and composition and a monumental firmness’, was being written.Footnote ¹⁴ The history of musical notation in Anglo-Saxon England in the tenth century provides one striking example of how the study of musical notations may open windows onto a wider perspective of the history of writing and of the book in Anglo-Saxon England: and, moreover, could be taken to a point of further precision which would allow identification of origin/dating of specific examples.

This catalogue ‘of manuscripts written or owned in England up to 1200 containing music’ includes 364 entries: of these about fifty-five are for complete (or substantially complete) manuscripts, and the remainder are either for fragments (rarely more than a few leaves) or for very short additions. A number of entries are for material from one book now surviving in multiple locations – such as a missal written at Exeter in the mid-eleventh century of which parts of eleven leaves are now in libraries at Lincoln, London and Oxford [H203]. The number of individual manuscript sources represented in the catalogue is thus effectively closer to 400.

The arrangement of individual entries is carefully explained. Nearly all have an introductory paragraph, and for complete manuscripts this is followed by an account of the content of the manuscript, each item numbered in one sequence. Following this there may be as many as four paragraphs: the first characterises the script of the manuscript and may identify the number of text scribes; the second characterises the notation(s) found in the manuscript, and usually identifies the number of music scribes, but does not always discuss whether one or more of the text scribes can be identified with one or more of the music scribes; the third supplies a basic physical description, usually including the number of leaves, their size, collation (only for manuscripts described in full) and the size of the written space and the number of lines, and sometimes observations concerning the ruling, retrimming and other matters. The final paragraph mentions origin and provenance (if known), followed by highly selective bibliography. This format is more or less the same for manuscripts which may have only a little notation – where often only the items with notation are described, and for manuscript fragments. The catalogue is followed by a bibliography, an index of incipits, an index of subjects, and eight plates.

The nature of the catalogue is declared clearly in its title, echoing Helmut Gneuss's Handlist of Anglo-Saxon Manuscripts: A List of Manuscripts and Manuscript Fragments Written or Owned in England up to 1100.Footnote ¹⁵ However, its inspiration is explicitly acknowledged to be the Catalogue of Manuscripts Containing Anglo-Saxon by N. R. Ker,Footnote ¹⁶ and the arrangement of its entries closely follows this model. In the Preface, the author explains how, having ‘begun to gather references to liturgical manuscripts of English origin dating roughly before 1200’, he came ‘to realize that in spite of’ catalogues in the British Library, Frere's Bibliotheca Musico-Liturgica and Van Dijk's typescript catalogue in the Bodleian Library, ‘students were still left with the task of assembling their own moderately comprehensive assessment of even one liturgical book’, and immediately adds: ‘Something closer to Neil Ker's Catalogue was required.’Footnote ¹⁷ This primary focus on liturgical material rather than on purely musical material, including notation, helps to explain the design of the catalogue entries, and to a large extent, a certain lop-sidedness in treatment and choice of material. Books with restricted, sometimes minimal, notated content may be catalogued in considerable detail: for example, the content of each section in the eleventh-century Benedictional London, British Library Harley 2892 [H158] is fully treated, with a list of the headings for every benediction, the whole entry taking up five pages. The musical content of the book consists of three notated incipits, and an antiphon added in the second half of the twelfth century. The text of this book had been edited in 1917.Footnote ¹⁸ In contrast, a late eleventh-century book of enormous musical interest – containing one of the earliest extant copies of Hucbald's De harmonica institutione as well as a large collection of songs, some with musical notation – Cambridge, University Library Gg.v.35 [H13], is treated cursorily. There is no list of the separate contents of the manuscript: rather this catalogue lists only texts with which neumes are found, the whole entry taking one and a half pages in total.Footnote ¹⁹ The user of the catalogue will need to look elsewhere to get a proper sense of the content, shape and form of this complex book. Hartzell explains his decision ‘to parade the contents of each pontifical or benedictional containing music in its entirety’: as the ‘largest single class of musico-liturgical books surviving from the period’ it was important to ‘reveal their idiosyncrasies’, ‘so that liturgical affinity with other sources, if it exists’ could be ‘determined easily’.Footnote ²⁰ That was a decision of great significance, since the result allows the chants which are in this case the primary object of the enterprise (whether notated or not) to be seen in their proper liturgical context, rather than divorced from it. Of course, the same would have been true of the seemingly miscellaneous hymns and songs listed in the entry for Cambridge Gg.v.35.

This unevenness in treatment extends to the choice of sources to be included in the catalogue. On the one hand, and despite the book's title, at least thirty-four sources which were never in England before 1200 are listed and analysed:Footnote ²¹ examples, written up at some length, include the Irish Gradual Oxford, Bodleian Library Rawlinson C.892 [H285] and the Norman Troper London, British Library Royal 8 C.xiii [H181]. On the other hand, sources of considerable significance, and certainly known through publications, remain unmentioned: examples from either end of the chronological range include the Codex Amiatinus (Florence, Biblioteca Medicea Laurenziana Amiatino 1), a Bible copied in Wearmouth-Jarrow in the late seventh or early eighth century (with quite a few neumed additions), and London, British Library Burney 357, which contains, notated in the second half of the twelfth century, a setting in two voices of the sequence Amor patris et filii.Footnote ²² Even if the origin of Burney 357 is unclear (the British Library website says ‘England or France’), its medieval provenance by 1200 was the Cistercian Abbey of St Mary's, Thame.Footnote ²³

In many cases, the nature of the material dealt with is such that to identify it and then to decide whether or not it deserved inclusion must surely have presented challenges to the author: often, notation may be present in the form of one or two neumes in a margin, or a pen trial on an endleaf, or, in the case of a group of St Albans books, over two words in a calendar.Footnote ²⁴ Few cataloguers mention such entries, yet their presence can have interest far beyond the world of musical history. The gospel book New York, Public Library De Ricci 115 [H208] was made in Brittany in the last third of the ninth century; notation for one line (Christ's words on the cross in the Matthew Passion) might already have been entered before the book was brought to England.Footnote ²⁵ As in the case of Pembroke 46, however, the notation is of interest whether or not it was written in England. That book was included in the Catalogue. London, British Library Cotton Claudius B.v, a book made east of the Rhine in the second quarter of the ninth century, and containing the Acts of the Council of Constantinople (680), was acquired by King Æthelstan, and given by him to Bath Abbey (probably in the 930s).Footnote ²⁶ Bischoff's 2004 catalogue entry for this book notes ‘Neumenfederproben’, which, whether German or Anglo-Saxon, would be of certain interest, and contribute to understanding the use and movement of this book.Footnote ²⁷ This book is not included.

Inconsistencies in the inclusion or not of such material can surely be forgiven; other omissions are less excusable. The most prominent Anglo-Saxon book not present in the catalogue is a handsome copy of Boethius' De consolatione philosophiae, probably made at Canterbury in the late tenth or early eleventh century. Sold at Sotheby's in July 2005, this book was for many years in the collection of Martin Bodmer, catalogued as MS 175.Footnote ²⁸ Among early medieval copies of the De consolatione philosophiae, this is probably the most heavily notated of all, with notations for fourteen (sic) of the metra; several different hands wrote the notations, with one dominating the work.Footnote ²⁹ Two more sources of Boethius' text, both associated with Christ Church, have also been omitted (Paris, Bibliothèque nationale de France lat. 14380 and lat. 17814): both have some neumatic notations for the metra, even if much less than in the Bodmer copy. This neglect of non-liturgical songs is reflected in the additional omission of London, British Library Royal 15 B.xix and Paris, Bibliothèque nationale de France lat. 4839 (with notation for some passages of Priscian);Footnote ³⁰ Oxford, Oriel College 3 (with notation for Prudentius' hymns); and London, British Library Cotton Julius A.vi (with notation for two medieval verses).

It is evident that some of this material was left out because the enterprise of visiting ‘all the repositories listed in Gneuss's Handlist proved impossible’;Footnote ³¹ Hartzell did not wish to include material without seeing it in person,Footnote ³² and as he puts it ‘a halt had to be called, and I have called it’.Footnote ³³ Nonetheless, it would have been useful to have a simple one- or two-line entry to represent those sources with musical notation of which he was aware but had not been able to see directly (presumably the Bodmer Boethius was one such): one of the principal features of such a catalogue as this ought to be inclusivity – the ability to more or less comprehend the whole field of surviving pre-1200 material containing notations. Equally, for manuscripts seen by the author but rejected because they did not satisfy his criteria for inclusion,Footnote ³⁴ a procedure whereby the fact of consideration followed by rejection is noted (such as was employed in Ker's Medieval Libraries and Bischoff's Katalog) would have been extremely welcome.Footnote ³⁵ In Appendix 1 below, we list some sources which we believe should have been included. In Appendix 2 we list others which, on the grounds of their origin and/or provenance up to 1200, we would exclude from the catalogue.

By the late tenth century musical notation was evidently in wide use in England: of the 364 books represented by the catalogue, fifteen are extensively notated books (such as Graduals and books including chants for the Divine Office),Footnote ³⁶ preserved more or less complete; and about forty-five are books with less musical notation (such as Pontificals and Saints' Lives with offices), also preserved more or less complete. A further 160+ are fragments from books of those two categories. It is the other 140+, including books into which musical entries were made, as supplements of various kinds (such as the neumed metra of Boethius) or simple pen trials, which underline the relatively common ability to read and to write musical notations, above all from the early eleventh century on; of course, once notation on lines became prevalent (from the beginning of the twelfth century in England), the business of writing notation in any form supplementary to what was intended by the scribe who laid out textual material was inhibited. Ruled staves required more space and therefore more planning than neumes, which could easily be fitted in between text lines, even those with glosses already massed around individual words.Footnote ³⁷

The widespread use of musical notation in England from the late tenth century on is also indicated by the variety of ways in which it could be written: Breton or Anglo-Saxon neumes in this first period, Anglo-Saxon neumes dominant for the greater part of the eleventh century, with a few examples of French or German notations evidently written by visitors,Footnote ³⁸ and then, increasingly towards the end of the century, ‘Anglo-Norman’ forms, often more diastematic (that is, with elements which clarify pitch). A new chapter in the history of notation begins with the use of dry-point or inked staves, onto which neumes were written. By the end of the twelfth century the earlier neumes had been transformed through the need to mark a point in space precisely: this determined the thickening of a previously insignificant script element (the top of a neume stroke) into a more prominent point, and eventually to solid ‘squareness’ – while the originally significant element (the stroke) became the now much less important tail attached to the point/square note.

Descriptions of notation provided in this catalogue use the distinctions ‘Breton’, ‘Anglo-Saxon’, ‘French’, ‘German’, ‘Messine’ and ‘Anglo-Norman’ neumes; for Anglo-Saxon neumes the detail of formation of individual neumes is often described, as are also the axes of rising and falling strokes – often a crucial feature in the identification of script-type and specific hand. For notations of the later eleventh century there is often mention of ‘diastematy’, and for notations on lines there is usually a concise description of the number of staves and manner of their preparation.Footnote ³⁹ Although Anglo-Saxon neume forms are addressed in the Introduction (pp. xxi–xxiii), there are no diagrams or small reproductions to guide those unfamiliar with music scripts from verbal descriptions to identification of different neume forms.Footnote ⁴⁰ And, while the Introduction provides references to items in the catalogue, a user who does not have direct access to those sources or to reproductions of them will be none the wiser. Overall the policy adopted here of not making routine reference to reproductions is very much to be regretted. Even if in such a catalogue it is not possible to include more than a small number of pictures (there are eight at the back, not cross-referenced to the Introduction), the value of the book as a guide to notation in England in this period would have been immeasurably increased by such references.

In cataloguing manuscripts with Old English, Neil Ker provided in an introduction a series of succinct observations and summaries drawn from the material he catalogued, including sections on palaeography, decoration, medieval and post-medieval provenances and even a section on ‘scribes and scriptoria’.Footnote ⁴¹ Ker also included a ‘palaeographical and historical index’ and an ‘index of owners’.Footnote ⁴² The lack of such explanatory material and indexes in the present volume is indicative of a deeper fracture between this and its OE model. For, beyond classification of notational type at a fairly basic level, descriptions of notation are relatively undisciplined, and made without reference to any theorized pattern: thus, for example, forms of neumes written on lines which are conspicuously different from those not written on lines, but have not yet reached a point at which they could be described as ‘square’ can be described in many ways (‘Anglo-Norman point neums’ [H305],Footnote ⁴³ ‘early square notation’ [H68], ‘incipient square notation’ [H69 and 353], ‘squarish notation’ [H73], ‘rudimentary square notation’ [H154], ‘notation well on the way to square forms’ [H233], ‘beautifully executed, almost square notation’ [H352]). Facing such a variety, the user of the catalogue will not easily make progress in determining questions of difference between notations, the chronology of their change, and the origin and dating of different sources. That all these descriptions are heading in the same direction is clear, however. At another level, in the descriptions of purely neumatic notations, this relative lack of discipline can be much more confusing: for example, what can it mean to describe three lines of notation as ‘Early Anglo-Saxon neums (or their close relative)’ [H119]? Or the following series: ‘rudimentary mixed neums’ [H133], ‘a mixture of a number of elements, mostly Breton’ [H14], ‘a hybrid of French and German forms’ [H277], ‘Anglo-Saxon neums or their sibling’ [H141]?

Yet there is a real problem in current scholarship which underlies these uneasy attempts at characterization. Most studies of neume types derive their examples from situations and periods by which a certain stability had been achieved – through the existence of a strong monastic institution with a set style (of which St Gallen is the most obvious example) or of a strong shared regional practice (as in the Beneventan region of southern Italy). Between the arrival in any one region of the idea of musical notation (and models for writing it), and the local development of classic, ‘school’ scripts, there is likely to have been a diversity of notations, some highly sophisticated in the amount of information they convey and others fairly basic. In this ‘in-between’ stage in the development of scripts (not to be confused with the development of notational techniques to convey information), the work of individual scribes in one locality might differ in the axes of script, in the ways of forming individual neumes and in the way in which those neumes were written in relation to each other and to the text on the page. The gap between the reality of these relatively early scripts and the neume tables printed in musicological studies based on ‘school’ scripts may sometimes be wide, with the result that a cataloguer may arrive at such a description as ‘rudimentary mixed neumes’ or ‘early Anglo-Saxon neumes (or their close relative)’ when in fact, leaving aside teleology, the way in which the music scribe was working was more fundamentally conceptual than the deliberate mixing of neumes derived from different local practices.Footnote ⁴⁴

There are several basic questions that any catalogue needs to have addressed, in particular how reliable it is concerning matters of date, origin and physical state of the manuscript. Generally, the dates given to the items in this catalogue appear to be sound, with occasional surprises. For H14, a musical entry in the margins of an eighth-century Gospel book (Cambridge, University Library Kk.I.24), the catalogue has ‘s.ix’, a date that is plainly too early for a rather clumsy Anglo-Saxon hand (which belongs somewhere in the tenth century). Of course, a significant number of the items can be said to be well known, and Hartzell had a good deal of help from earlier literature in determining dates, but our examination of some of the previously unpublished fragments shows that he can usually be trusted.

In the case of one particular group of sources, however, the importance of the correct dating of the text scripts is paramount, because they are linked with the earliest use of stave-lines in England. It is odd that this is not even touched upon in the introduction, and to determine from the text of the book when this took place is a useful illustration of its strengths and weaknesses in the treatment of manuscript and printed sources. The headnote to British Library, Harley 3908 [H164] suggests that the following office ‘may be the earliest English witness to the adoption of musical notation on a four-line staff’, dating from ‘s.xi/xii’.Footnote ⁴⁵ Unfortunately, this does not stand up to scrutiny. The office was written in mature Christ Church script (a late form of Caroline minuscule developed in the late eleventh century and written at Canterbury and Rochester). Hartzell followed the date given to the office in a 1991 article without noting that the 1999 handlist by Gameson gave it the broad date of first half of the twelfth century.Footnote ⁴⁶ The script looks too early for c. 1100, and, as the distinctive scribe is identifiable in two manuscripts almost certainly written in or about the 1120s at St Augustine's, Canterbury,Footnote ⁴⁷ the suggestion that Harley 3908 has an early use of staves is certainly true, but they are not as early as the catalogue suggests.

The fragment Cambridge, University Library Inc. 5 B.3.97 [H18] is dated ‘s.xi ex.’ and said to be noted on a ‘four-line drypoint staff’ with ‘two lines of each staff … coloured green’.Footnote ⁴⁸ A close examination suggests that the stave-lines were coloured alternately green and yellow, the yellow having largely faded or been abraded so as to make it very difficult to see. (Staves in green appear to occur in only one other entry, Oxford, Bodleian Library Rawlinson C.892 [H285], an Irish-made book of the mid-twelfth century, and the only other stave-lines in a colour other than red occur in Oxford, Bodleian Library Selden supra 90 [H290], described in this Catalogue as in orange.) Cambridge Inc. 5 B.3.97, as Hartzell observed, was ‘written in a compact, angular Norman hand’, but whether as early as the end of the eleventh century seems doubtful. Dating these kinds of small hands is certainly not easy, but a safer date might be 1100 × 1130, and, as this item is a fragment with no evidence of an early English provenance (although the catalogue states ‘written in England’),Footnote ⁴⁹ its scribal evidence deserves to be treated cautiously. Cambridge, University Library Add. 2748 [H4] is dated ‘s.xii in.’, and described as noted on a ‘Drypoint staff of four lines’.Footnote ⁵⁰ This is another fragment with no evidence of an early English provenance (although also stated to be ‘written in England’). The nationality of the text scribe is not characterised by Hartzell, but this was also written by a Continental scribe, and its evidence also deserves to be treated cautiously.

London, British Library Royal 12 B.xxii [H182] and London, Lambeth Palace 359 [H198], both dated here ‘s.xi/xii’, have dry-point stave lines. Hartzell notes that the origin of Lambeth 359 ‘is uncertain’, although the noted additions were ‘written in Normandy’ (presumably on the strength of the ‘excellent Norman notation’), and properly adds that ‘there is no indication when it might have come to England’.Footnote ⁵¹ M. R. James suggested the manuscript might be Italian or Spanish,Footnote ⁵² and it is not admitted in either Gneuss or Gameson.Footnote ⁵³ We would tentatively suggest, on the strength of the author portrait on fol. 12, an origin in or near Brittany. Taking all this into account, it would have been wiser not to have included the manuscript in the catalogue at all. The same could be said of Royal 12 B.xxii, for this is correctly said to have been written in France, and there is no evidence that it was early in England. The notation on lines in Salisbury Cathedral 25 [H322] is for a Marian sequence on the last verso, and Hartzell cites the study on the cathedral manuscripts by Teresa Webber for its date.Footnote ⁵⁴ However, the sequence is not mentioned by Webber in her careful scribal analysis of the manuscript, and this suggests that she considered it to be an ‘addition’. A brief note of explanation could have been included by Hartzell to clarify the situation, and, at present, the early date given to this notation on staves (‘s.xi/xii’) should also be treated cautiously.

The earliest certain use of staves in England dates from or a little before the 1120s. Cambridge, Corpus Christi College 371 [H41] comprises the openings of three hymns noted on four-line dry-point staves, the texts of which were written by Eadmer of Canterbury. As sometime precentor of Christ Church, Eadmer would certainly have been able to read music, and as he is known to have been a good scribe it is likely that he could write it as well.Footnote ⁵⁵ The hymns occur in a manuscript of Eadmer's theological works almost entirely in his own hand (Hartzell has ‘probably’, which is unnecessarily cautious), and it appears likely to us that the scribe of the text is identifiable as the scribe of the music. (It is not clear from Hartzell's description whether he considers the text and music scribe to be the same or not.) That is to say, Eadmer wrote the text and notated the hymns, using staves, almost certainly later than 1121 and before his death sometime in the late 1120s. Running him a close second would be the scribe of Eton College 80 [H99] who wrote and notated (on four dry-point-ruled lines) two short pieces near the end of a manuscript containing a work by Jerome. Hartzell follows the date given to the manuscript by N. R. Ker of ‘s.xii in.’,Footnote ⁵⁶ but this could have been refined a little. He did not notice that the scribe of the manuscript is the best-known Rochester scribe of the early twelfth century, responsible for the important Textus Roffensis (still at the cathedral), datable to 1122/3.Footnote ⁵⁷ A very good and accurate scribe, he was probably active somewhere between about 1105 and 1130, and as Eton 80 is included in the catalogue of Rochester books of the 1120s in the Textus Roffensis, his use of staves may be a few years earlier than their use by Eadmer in Corpus 371. There are other examples in the Catalogue of musical notation on lines dated by Hartzell between ‘s.xi/xii’ and ‘s.xii (1st quarter)’, but we have not yet studied these, and the dates given to them might well be correct, in which case what is said above will need to be amended.Footnote ⁵⁸ However, this matter does demonstrate that the sources dated in the Catalogue between about 1070 and 1130 are those which we suggest should be treated with most caution. While Hartzell must be commended for drawing attention (for the first time) to a large body of material of this date, these sources belong to a period of considerable variability and instability in English handwriting.Footnote ⁵⁹

In 1919 M. R. James wrote ‘see books for yourself; do not trust that the cataloguer has told you everything’.Footnote ⁶⁰ This is a sentiment that can be applied to almost all catalogues and the present catalogue is no exception. The physical descriptions provided are clear and reasonably thorough, but there are occasions when, to have a satisfactory sense of the relation between physical make-up and the contents of a manuscript, the reader needs to have fuller information. The Pontifical Cambridge, Corpus Christi College 146 [H24] is a complicated book, made at an unidentified centre: the main core of the book is written in a hand of a type firmly associated with Canterbury, yet the content of certain entries refers to the Old Minster, Winchester (which could, of course, depend on an exemplar).Footnote ⁶¹ The book was later extensively augmented at Worcester. In its present state the book is constructed from six parts that can be distinguished physically and scribally, as follows:Footnote ⁶²

Endleaves: fol. i and pp. 1–6

Part 1: pp. 7–22 + 31–52

Part 2: pp. 23–30 (a near-contemporary addition to Part 1)

Part 3: pp. 53–60

Part 4: pp. 61–318

Part 5: pp. 319–22

Part 6: pp. 323–8 (these once followed p. 280 in Part 4)

Endleaves: pp. 329–30 and fol. ii

Hartzell is well aware of the composite nature of the manuscript, but does not share much of the available information with the reader: his remarks on the make-up of the book (he divides it into three parts, corresponding to the above analysis as HI = up to the end of Part 3; HII = Part 4; HIII = Part 5 onwards) are set at the front of the description, while at the end he refers to ‘a distinctive, elegant, generally upright Caroline minuscule by a single scribe’. Attribution to a single scribe is patently misleading, and must here refer to the core of the older book, fols. 61–318, certainly written by one highly distinctive scribe.

One other example where brevity has led to ambiguity concerns a manuscript of such interest to historians of music and liturgy that it deserves clarification. In the later Winchester troper, Oxford, Bodleian Library Bodley 775 [H263], a cycle of Alleluias for the liturgical year is copied on fols. 76^r–87^v. Such cycles are usually made up of constituent parts, a Temporale section, stretching from the beginning of Advent to the feast of Pentecost, then the Sundays after Pentecost, and a Sanctorale. Except that in this cycle the post-Pentecost series is broken into parts, and mixed up with the Alleluias for saints' feasts. Realising this, Hartzell decided to print the Alleluia series in a corrected liturgical order, that is, not in the order as it stands in the manuscript; his listing is provided with folio numbers, so that it is entirely clear when it deviates from the actual state of affairs. Before embarking on the list he provided the explanation: ‘the correct sequence of folios is 76–80^v, 85–86^v, 83–84^v, 81–82^v’.Footnote ⁶³ That sequence certainly presents the correct order of pages in terms of the content, but it is highly deceptive in relation to codicology: despite what the folio numbers might imply, 85–6 is not a bifolium, nor is 83–4, nor is 81–2 (nor do we believe that Hartzell intended his explanation to be read in this way, since he was evidently aware of the correct collation).Footnote ⁶⁴ What had happened is this: the gathering now represented by fols. 80–7 was worked on by the text scribe, who wrote out the Alleluia cycle in a clear order – Temporale, Sundays after Pentecost, Sanctorale – exactly as represented in this catalogue's listing of them.Footnote ⁶⁵ Between the work of the text scribe and that of the music scribe, however, the two bifolia between the outermost and innermost in this gathering got reversed:Footnote ⁶⁶

The order of folios as prepared by the text scribe:

• 80 ————|———— 87

• 85 ————|———— 82

• 86 ————|———— 81

• 83 ————|———— 84

The order of folios as notated by the music scribe:

• 80 ————|———— 87

• 81 ————|———— 86

• 82 ————|———— 85

• 83 ————|———— 84

The music scribe was then confronted by various situations where the join between the end of a verso and the beginning of the now following recto did not quite fit: at the join between fols. 80^v and 81^r the word ‘Alleluia’ at the end of the verso was followed at the beginning of the recto by ‘Alleluia’ again; at the join between fols. 84^v and 85^r the word ‘Alleluia’ before a verse was missing and needed to be added. In the first case he simply did not enter notation over the extra (first) ‘Alleluia’ (at the bottom of fol. 80^v), and in the second case, he added the required word ‘Alleluia’ and notated it (at the bottom of fol. 84^v). In the case of the other two unexpected joins (between fols. 86^v and 87^r, and between fols. 82^v and 83^r), the placing of ‘Alleluia’ at the bottom of the verso, followed by a verse on the following recto, allowed the music scribe to proceed without the need for textual emendation.

Carefully made descriptions of manuscripts may sometimes lead a cataloguer to miss the connections or links that can tie one manuscript to another, and it is often these that begin to make a book rather more than an isolated object. In this Catalogue, the severe limitation of bibliographical references is matched by a similar constraint in relation to known links between manuscripts. Yet often, detailed examination of scribal histories may throw light on times and dates of copying, and not least, how the notation of specific examples fits in with the notation of other books. Examples which illustrate the significance and complexity of the challenge are offered by three books and one fragment, all almost certainly made in Canterbury, for two of which the contents are described here at some length. London, British Library Harley 2892 [H158] is a Benedictional, written by a good (text) scribe in a late English Caroline script,Footnote ⁶⁷ perhaps, as its content would suggest, at or for Christ Church.Footnote ⁶⁸ (In this Hartzell went beyond Ker, who did not include the manuscript in his list of books linked with Christ Church in Medieval Libraries of Great Britain.Footnote ⁶⁹) Hartzell dates the Harley manuscript to s.xi 2/4, in slight contrast to Ker's dating of another source, which he suggested was probably copied by the same scribe, as s.xi med (Cambridge, Trinity College B.15.34).Footnote ⁷⁰ Cambridge, Corpus Christi College 44 [H22], a very fine Pontifical, is written in a monumental late English Caroline minuscule. Hartzell dates this manuscript as s.xi 3/4,Footnote ⁷¹ and states at the beginning of the entry ‘pontifical of Christ Church’, but at the end ‘written at St Augustine's, Canterbury’. In fact, this book was included in Ker's Medieval Libraries under St Augustine's,Footnote ⁷² probably because of the identification by T. A. M. Bishop of a slightly later scribe who wrote interlinear glosses both here and in other books originating at St Augustine.Footnote ⁷³ Bishop also identified the main scribe of Corpus 44 as the scribe of a fragment at Canterbury Cathedral (Add. 20), containing the Rule of Chrodegang.Footnote ⁷⁴ Although he did not record Bishop's identification, Ker suggested that this fragment might be from a Christ Church book.Footnote ⁷⁵

The confusion of these perspectives on the relation of three books and one fragment to the two Canterbury houses is plain – and the lack of a solution to the puzzles raised by their scribal interrelationships is hardly to be set at the feet of an author trying to provide a catalogue of a broader selection of material. Nonetheless, uncertainties in the literature concerning the origin and provenance of the books that Hartzell catalogued could have been directly mentioned. And can the Anglo-Saxon neumes entered for the Ordinatio episcopi in Corpus 44Footnote ⁷⁶ be related in any way to the musical entries in Harley 2892? The notation of books made in Canterbury is at one and the same time both the most unstudied and the most important outstanding issue in the history of early musical notation in England: therefore the relation of musical entries in two books copied by scribes obviously working in the same period is worth consideration (even if neither book was originally prepared for musical notation).

Making a new catalogue of any kind is a brave undertaking, since it offers the opportunity to re-evaluate everything that is included. A good example is London, British Library Cotton Caligula A.xiv, fols. 1–36, the famous ‘Cotton Troper’ [H128], a manuscript best known for its exotic pictures that have long been agreed, with one exception (a Gospel-Lectionary, Cambridge, Pembroke College 302), to have no parallels, either in late Anglo-Saxon manuscript painting or in contemporary Continental painting.Footnote ⁷⁷ The repertory of tropes copied in this book is most closely linked to that of the two Winchester Tropers,Footnote ⁷⁸ and the text scribe has usually been assumed to be English, writing a late English Caroline script. Hartzell appears to be neutral, stating that the manuscript was written in a ‘compressed, formal text hand’,Footnote ⁷⁹ but whether this was deliberate is uncertain. However, he does characterise the notation as ‘Anglo-Saxon’ and as the work of the ‘main text scribe’, so the implication appears to be that the scribe was English. Taking into account the work, and the many reproductions, that are now available on English Caroline minuscule, it now seems clear that the scribe of the manuscript is Continental. (We suspect that the scribe has been thought to be English for so long that his script has always been regarded in an unquestioning fashion as ‘English’.) The general aspect of the hand is a little loose and fluid, and this is simply not a characteristic of English-trained scribes, whose work, even in poor examples, is usually firmer; no certain English hand like this one has been successfully identified in the English sources. Many of the details are also not English (some of which Hartzell notices), such as the use of long r, the small, angular headstroke of the a, the form of the g, and the variable finishes to the final minim of the n and m. The degree to which the scribe was influenced by late English Caroline is a matter that remains to be determined, but there are aspects of the notation which could set it outside England also, above all, the pointed tops of the clives – a standard characteristic of many French neumatic scripts in the second half of the eleventh century, and relatively rare in Anglo-Saxon examples. (In Anglo-Saxon neumatic scripts written at both Canterbury and Winchester before the Conquest a pronounced roundness at the top of this stroke was cultivated. That roundness remains characteristic, even if not so exaggerated, in the many later examples of notation made at Exeter and Worcester.) The closest substantial example of neumatic notation written in England with something approaching a point at the top of the clivis stroke is in Cambridge, Corpus Christi College 267 [H33 and Pl. V], which has an office for St Mellitus added on flyleaves of a book made at St Augustine's: even here the point seems to result more from the contrast of thickness and thinness as the pen is turned, than from the more deliberate stopping and change of direction as the highest point is reached, as in the work of the Cotton Caligula music scribe.

What this and other examples discussed above demonstrate is the uncertainties that still surround many manuscripts in the catalogue. Explanatory notes, with reference to bibliography that could contribute to an understanding of the conditions and circumstances in which individual manuscripts were produced, are very limited, leading to an impression of confidence about matters concerning script, origin, provenance and date that may mislead the unwary reader.

Although the book is designed with care, with one important exception it has not been made easy to use. There are two substantial indexes, one of incipits and one of subjects, both of which are directed towards readers who are concerned with liturgical texts and subjects. There is no listing of the catalogued items by type of book (Gradual, Pontifical and so on), by date, by origin (whether by country or by centre), by medieval provenance (that is to say ownership), by language (there are several Anglo-Norman or French texts, and some in Middle English), or by type of notation. The absence of indexes or lists for matters other than the directly liturgical means that to find out anything concerning the manuscripts other than their content, it will be necessary to look through the entire text. This can be nothing but time consuming. And whether all of the codicological or palaeographical information provided is always necessary is questionable. The description of the forms of some abbreviation and individual letters in many entries may be useful for pre-Conquest manuscripts, but very rarely for post-Conquest ones. Such descriptions are, in any case, only really useful when accompanied by reproductions, and of the eight items reproduced only two, perhaps three, are pre-Conquest. The eight plates show full pages arranged in order of their dates, and each of them was presumably chosen as representative or characteristic of their time and place. (The author gives no reason for his selection of the reproduced manuscripts.) The scale of reduction is not stated, and it is not always possible to work this out.Footnote ⁸⁰

Yet the making of this catalogue was driven from the beginning by a wish to encompass evidence of liturgical practice in England, especially liturgical items which were sung (to more than a simple recitation formula). On this level the catalogue is a major success: the quality of liturgical analysis is excellent (and, as far as we can determine, free from error) and the clarity with which liturgical items are edited and displayed likely to be helpful to all, from novice to specialist. The two indexes support this aspect of the enterprise admirably. The absence of Proper trope incipits in the descriptions of Corpus 473, Bodley 775 and Cotton Caligula A.xiv is to be regretted: it has the odd result that chants only represented by incipits in these three books are listed in the catalogue, whereas the fully written out and notated tropes, the material around which those books were organized, are not. More seriously, for a small number of chant books and fragments of such books, only festal headings are listed, without (or with only partial) details of individual chants: this renders the index of chants relatively incomplete. Nevertheless, that lack does not seriously detract from the considerable value of this catalogue as a liturgical handbook, a most useful guide to liturgico-musical practice in the period up to 1200 as it can be recovered from extant manuscript sources.