John Williams

title.none: Walker, Views of Transition (Williams)

identifier.other: baj9928.9904.010 99.04.10

identifier.issn: 1096-746X

description.statementofresponsibility: John Williams, Pittsburgh University,

publisher.none: .

date.issued: 1999

identifier.citation: Walker, Rose. Views of Transition: Liturgy and Illumination in Medieval Spain. London: The British Library and The University of Toronto Press, 1998. Pp. 264. $75.00. ISBN: 0-802-04368-2.

type.none: Review

relation.ispartof: The Medieval Review

The Medieval Review 99.04.10

Walker, Rose. Views of Transition: Liturgy and Illumination in Medieval Spain. London: The British Library and The University of Toronto Press, 1998. Pp. 264. $75.00. ISBN: 0-802-04368-2.

Reviewed by:

John Williams
Pittsburgh University

This book looks at the effects, measured in terms of the texts and illumination of liturgical manuscripts, of the conversion in Spain around 1080 from the ancient Hispanic liturgy, called Visigothic or Mozarabic, to the so-called Roman one practiced elsewhere in Western Europe. As the author states, "Liturgical change will be studied through close analysis of text and image in manuscripts of the Mozarabic and Roman liturgies produced in Spain before, during and after the change in liturgy. The first part of this book (Chapters 1-3) will deal with manuscripts containing texts for the offices in which all monks could participate, while the second part (Chapters 4-6) will examine manuscripts containing texts for the mass, where the role of the celebrant will be paramount. The texts will be interrogated for detailed information on the form and structure of the change, its extent and application'.The presentation of the liturgical text will then be examined to determine the ways in which the change was expressed and made visible." (22) By "made visible" she means the illumination of liturgical manuscripts. In fact, the focus is on the northern kingdoms of Leon and Castile under Alfonso VI (Ý1109), Galicia excepted. And within this perimeter, in so far as it is possible, she attempts to employ the liturgical manuscripts of the Castilian monastery of Silos.

The Introduction continues with a presentation of the background of the liturgical change. It proceeded under Pope Gregory VII's initiative as part of his reform movement and was facilitated by the francophilic ruler of Leon-Castile, Alfonso VI, with the encouragement of Hugh of Cluny. Gregory maintained that Spain had been converted originally under the Roman liturgy when Seven Apostolic missionaries had been sent from Rome by Peter and Paul, but had subsequently lapsed. So the change was in fact a return. His attitude assumes, Walker says, that the old liturgy was tainted by heresy, a judgment linked by her to the Adoptionist controversy of the late eighth century despite the fact that Gregory's concern about Spanish heretical signs was voiced a decade earlier than Adoptionism. The difficult task of assessing opposition to the change is attempted, and she usefully provides in Appendix I a translation of the defense of the Mozarabic liturgy penned at the Aragonese monastery of San Millan de la Cogolla. At the center of change she places the Leonese abbey of Sahagún so favored by Alfonso VI, where two successive Cluniac monks were installed as abbots at the time change was introduced. In Chapter One the examination of manuscripts begins. As befits a book published in a series sponsored by the British Library, the manuscripts chosen as specimens are, to the extent they can be, there. For the Mozarabic office they are B.L. Add. MS 30844, Add. MS 30845 and Add. MS 30846; for the Roman rite liturgical equivalents she further selects from the Silos corpus Add. MS 30847, Add. MS 30848 and Add. MS 30849. All belong to the corpus of manuscripts once in the Castilian monastery of Santo Domingo de Silos. All but the last, which is written in Caroline, or Protogothic as she terms it, are written in the traditional, "Visigothic" peninsular script that would be replaced by Caroline in the course of the twelfth century. Walker holds that all of these were the products of Silos. She acknowledges that in a doctoral dissertation written by Ann Boylan, "Manuscript Illumination at Santo Domingo de Silos Xth to XIIth Centuries" (University of Pittsburgh, 1990) these manuscripts were assigned to other scriptoria. Boylan proposed that the scriptorium of Silos came into being to produce manuscripts for the Roman liturgy. Walker prefers to see a longer history for the Silos scriptorium and accepts all of her specimens as Silos manuscripts. In any case, she acknowledges that since Boylan's scriptoria are within a 50 km radius, it is not a crucial issue (50-52), although in her conclusion she gives this difference in opinion on the history of the Silos scriptorium some weight. (219-20). (It perhaps should be noted that I was the director of Boylan's dissertation. The only part that has been published is Ann Boylan, "The Library at Santo Domingo at Silos and its Catalogues (XIth-XVIIIth Centuries," Revue Mabillon, n.s. 3, v. 64 (1992) 59-102.) Walker subjects her manuscripts to a codicological analysis (52f.) covering dimensions and other aspects of physical character such as ruling, pricking, and abbreviations, script, liturgical abbreviations, orthography. From this she concludes that there is so little consistency in production that it "is not possible to identify a 'house' style." (56-7).

Chapter Two (69-101) compares the new organization of the liturgical texts to the old. The Mozarabic manuscripts placed monastic offices and masses together under the heading for the feast day in one volume. So the texts are normally ordered by liturgical day, under which vespers, matins and lauds and then the masses are given. Hence the label of "mixed book" (misticus). The Roman custom, on the contrary, was to have separate books for the mass (missal) and the hours (breviary). For reasons of room, therefore, the Mozarabic liturgical year could not be presented in a single volume. There is no standard format, but an idiosyncratic and thematic character. (69-72) Moreover, in the Mozarabic manuscripts there is no clear-cut division between the temporal and the sanctoral cycles. Although there was a "slight but perceptible move away from the commemoration of Spanish martyrs," very few of the saints included in the Mozarabic Calendar were excluded from the Roman. It is true, however, that the offices for them tended to be dropped. Add. MS 30848 and 30849 made up for this with many new "Roman" saints. (74-6). Appendix II provides a list of the contents of the manuscripts by heading. As a supplement to the comparison of the attitude to saints provided by the Appendix II, she treats the altars dedicated at Silos in 1088 as symptomatic (76f.) The central altar was dedicated to SS. Sebastian, Peter, and Andrew. Mary, Michael and John the Evangelist were in the altar in the right apsidiole, Martin, Benedict, Nicholas and Domingo in the left apsidiole. Of these, only the last three were "new".

A comparison of offices (80f.) reveals that the most obvious difference in the new liturgy is the growth of the night office of matins, mainly through the inclusion of readings. In the Mozarabic liturgy matins was centered on the singing of Psalms, but scriptural and homiletic readings were most prominent in the new texts. In Add. MS 30847 and 30848 most of these were authored by Smaragdus, although in the latter manuscript additions also include as readings for the four Sundays after the Easter octave excerpts from the Commentary on the Apocalypse by the eighth century Spanish abbot, Beatus of Liebana, a copy of which was written at Silos in 1091 (88f.). (So far as I know, with this revelation Walker has provided the only evidence for a liturgical use of the Beatus Commentary. We must be grateful for this discovery, but it is disappointing that the Beatus texts, which are not given, were not collated with one of the several published editions of the Commentary.) In the readings of the later manuscript, Add. MS 30849, another Carolingian reformer, Paul the Deacon, dominates, in imitation of French practice. In the use of rubrication (90f.) Walker finds light symbolism prominent and she suggests it was connected to a reform attitude.

In Chapter Three, "The Sacred Text Made Visual" (102-37), the manuscripts of the office are analyzed from the point of view of their decoration, with a generous selection of black and white illustrations and several color plates provided to help the reader follow the discussion. She acknowledges that at first glance they present a heterogeneous face, but she detects important similarities, chiefly the use of compartmentalized color, the hallmark of the Mozarabic school, and the use of the Vespertinum (VPR) monograms that introduce each office text. These monograms vary a great deal among themselves, being geometric, "tree-like" or, especially in Add. MS 30845, formed by human figures. The initials employ a variety of designs. Parallels with examples in manuscripts from Valeranica, San Millán and Toledo are signaled, but she finds that a similar system of decoration enforces a unity. These parallels might seem to be evidence for origins in these various scriptoria and for earlier dates than Walker proposes, but she rejects (119) the idea that they originated elsewhere than in the Silos scriptorium or earlier than some advanced time in the eleventh century (absolute chronology is not a focus). This conclusion receives an unacknowledged challenge when Walker turns (120f.) to the first of the manuscripts with the Roman rite, Add. MS 30847, for she describes the change in illumination as "dramatic" even though the script remains Visigothic. To begin with, she speculates, the Roman rite manuscripts are more sparsely decorated out of a desire to suppress Mozarabic ornament. Add. MS 30848 in turn differs strikingly from 30847 by employing "banner" headings, display texts of the style of the Beatus Commentary written at Silos in 1091. This is taken as an effort to lend this manuscript an explicitly antique signature, since this form of display originated in Castile in the tenth century. The last of the Roman rite manuscripts, Add. MS 30849, although written in Caroline script, uses many Mozarabic decorative features. Add. MS 30847 was an early stage in transition when the reader was led through the text with minimum distraction. Mozarabic decoration is almost entirely absent. Add. MS 30848 is "archaizing", forging a manuscript of a long Mozarabic lineage. Add. MS 30849 is essentially Roman in content and style. Some initials may have Mozarabic decorative motifs, but others continental ones. It probably dates from second quarter of the twelfth century, after which the Silos Scriptorium probably went into decline (134).

In the first chapter of Part Two, Walker selects the missals she will use to examine the processes by which the Roman mass texts replaced the Mozarabic. Missals are relatively rare, and it is necessary to employ a trio of manuscripts she does not claim are Silensian, although one of them, Salamanca University MS 2637, may be. It has a mass for Santo Domingo of Silos, but since it is not a special votive mass, the evidence is not strong. She dates it to the second third of the twelfth century. The Missal of San Facundo (Madrid, B.N. Vitr. 20-8) is another choice. It is from Sahagun, the primitive center of liturgical change. Internal evidence, namely a letter from Hugh of Cluny to Bernard de Sauvetat, the Cluniac monk chosen as abbot of Sahagun in 1080 in order to push forward with liturgical change, dates it to the approximate time that change was undertaken. A missal from the Aragonese monastery of San Millan de Cogolla (Madrid, Real Academia de la Historia Cod. 18) is the third example. It is dated by Walker to the last decade of the eleventh century. I have recently given my reasons for dating it slightly later (The Illustrated Beatus, vol. III, London, 1998, 25).

In Chapter Five (154-73) the contents of the missals are examined to "see the extent of uniformity between the Roman mass texts in view of the different origins of the manuscripts 'Then we will determine how different the Roman text is from the Mozarabic' " The extent of agreement between the Roman missals is shown to be "remarkable", and she recognizes "the gulf which separates them in form and content, if not in function, from the Mozarabic formulae." (167) One of the most significant differences between the Mozarabic mass texts and the Roman is found in the source of their lections. The Mozarabic is distinguished by 18 readings from the Apocalypse in the week after the Easter octave. A comparison of the readings for major feasts is provided in Appendix VI.

In the longest, sixth, chapter, "The Mass Displayed" (174-207), the respective systems of decoration in Roman and Mozarabic mass texts are compared. She finds in the Mozarabic books that "a constant rhythm of rubrication and minor initials with infill colours articulates the texts." Even if the style of ornament varied, "office and mass texts continued to be treated as a entity. There is a hierarchy of initials which operates in broadly the same way in each of the manuscripts." (174) From the separate consideration of the decoration of mass texts one perceives that "the mixed structure of the books is mirrored by the integrated decorative schemes. In total the mass texts receive more elaborate illumination than the office texts, largely because of the three major initials used to mark the lections. Otherwise the heading and vespertinum monogram used to begin the office balance the heading and initial used to begin the mass." (179) Walker then turns to the Roman missals. The decoration of the Facundus Missal and Aem. 18 differ sharply from the Mozarabic rite manuscripts. A lengthy description and comparison of these two Roman texts follows. Even so, they differ so significantly from each other visually that it strikes Walker that "it is difficult to believe we are dealing with the same ritual." (179) In fact, the same points in the text receive the major emphasis in each. The Facundus Missal from Sahagún opens with an initial P (per omnia secula) in the form of a man carrying a large P ornamented with a floral scroll and beast heads (pl. 7). On the verso of this first folio is a framed Vere Dignum initial V of floral interlace with beast heads (fig. 51). Smaller initials in this vein appear in the text (figs. 55, 56). The Te Igitur T takes the form of a drawing of a crucified Christ (fig. 52). In contrast, the P of the Per Omnia of Aem. 18 (pl. 8) is composed of a man holding a curving dragon, while the Vere Dignum monogram (pl. 8) is a design involving two inverted angels holding between them the Lamb within a mandorla. The Te Igitur of Aem. 18 displays an enlaced T and E against a framed panel. A single leaf added to the front of Add. MS 30849 provides still another example of the Per Omnia P composed of a man holding a dragon (fig. 45). Faced with these three examples of man-dragon initials, Walker proposes (184, 218, 221) that it stands for the defeat of heresy and the triumph of orthodoxy. Since the man and dragon initials employ different configurations, a verbal rather than visual model may have been involved. "If this is so, this common iconography suggests that there may have been more central control of liturgical change than has been assumed." (184) Moreover, "We might almost wonder whether the figure balancing the large letter 'P' with its dragon heads is not meant to be a visual witticism for Bernard's own demanding role in introducing the new liturgy." As this somewhat discursive chapter proceeds, Walker returns (192ff.) to the issue of the provenance of the decoration of the Facundus Missal. Despite the links with Moissac manuscripts indicated by Meyer Schapiro, Serafin Moralejo and myself (duly noted by Walker), she finally concludes, for reasons not clear, that it is "entirely plausible that a Burgundian manuscript, possibly sent through Cluny, was used as the model for the text [of the Facundus Missal]." (194) The ornament of Salamanca 2637 is then analyzed (196f.). It differs textually in structure and does not follow the decorative pattern established by the other two missals. It has a number of animal interlace initials (figs. 61-65) of indifferent quality. The grandest initial (fol. 2r), however, is splendidly carried out. It is an enframed A for the first letter of the opening of the first antiphon embedded in inhabited rinceaux (pl. 9). Walker proposes (199) that the illuminator was inspired by a lion capital in lower cloister at Silos (fig. 60), and finds this strengthens a Silos connection for the manuscript. However, since this is a classic example of the so-called Aquitainian rinceaux (Jean Porcher, French Miniatures from Illuminated Manuscripts, London, 1960 pl. XV), such an inspiration seems doubtful. After her survey of the illumination she concludes that, "If Vitr. 20-8 was the model and Aem. 18 is wavering reflection, Sal. 2637 is a new statement." (203) Moreover: The study of these three missals has established that the association of the new mass text with the new artistic style was very strong. This is reflected very clearly in the way the decoration articulates the text, in two of the manuscripts fixing the reader's attention above all on the preface and canon, which is in marked contrast to the use of illumination in the Mozarabic books. Moreover in the Roman missals there is no sign of the many 'Mozarabic' motifs which linger in the Roman breviaries; even Aem. 18 has only minimal 'Mozarabic' symptoms in its illumination as in its text. It appears from the illumination, as from the texts, that the liturgical change was uncompromising in the matter of the mass. It demanded a clear declaration of orthodoxy and strict conformity to the new order. (203-4) The Conclusion (208-24) provides an inclusive review. She speculates that since the new liturgy and the new style of illuminations may have appeared in the abbeys almost simultaneously, they may have come to signify one another (219). At the very end she comes to an unexpected and, for me, unjustified conclusion: "As a whole these manuscripts could be said to have provided a rather despondent picture of the process of change. There is scant evidence for enthusiastic acceptance of the new liturgy among those who produced the new manuscripts for many years after the initial edict."

This negative conclusion seems to be based on a dubious correlation between illumination and fervor. The presence of illumination in any text is problematic in that it is not strictly necessary to the function of the text. Gift or display texts meant to reflect glory on their owners or givers are an exception. But liturgical manuscripts are more functional in character, and as Walker herself notes at the beginning of her book, her manuscripts are not richly decorated. Her suggestion that monks may have come to associate a Romanesque artistic vocabulary with the liturgical change seems doubtful. To be sure, the new style of illumination was part of a current of Europeanization of which the change in liturgy played a part. But as Walker herself noted (218), the switch to imported, Romanesque, formulas occurred before the change of rite. She attempts to accommodate this fact with her desire to link liturgical change with Romanesque through a theory of subsequent estrangement between church and crown. In general, for me, Walker is too ready to assume some kind of equivalence in terms of prescription and orthodoxy between ornament and the text it appears with. There is insufficient recognition of the idiosyncrasies of individual scriptoria and scribes and circumstances. The role she assigns to ornamental vocabulary as signifier is to my mind exaggerated. Thus her remark upon noting the extent of difference between the illumination of the missals: "it is difficult to believe we are dealing with the same ritual." (179) The man-dragon formula that she proposes was read as the defeat of heresy was used for all kinds of texts. It was a stock item in the Romanesque vocabulary of forms. What guarantee do we have that when it appears in a liturgical manuscript it registers in a special way?

The issue of choosing decorative formulas for ideological reasons is raised again in her perception of Add. 30848, the manuscript that, as we reported above, employed a distinctive kind of "banner" display text: [The 'banner' heading] may represent an attempt to accommodate the change by denying it. By producing a Roman codex in a specifically antique Mozarabic style--and even including part of Beatus's long-established text--the scribe of this manuscript seems to construct a past which never was, as if the Roman liturgy had been used in tenth-century Spain'there is still no sign of the extravagant decorative style of the Silos Beatus'the design seems to have been very carefully chosen for its traditional nuances and for its purely geometric forms. (229) Aside from the premise that monks were historically sensitive to style, this statement is problematic in characterizing the décor of Add. MS 30848 as being more archaizing than that of the other Roman breviaries, and in citing its lack of figures as a sign of its archaic nature. As is well known, the text of the Silos Beatus was completed in 1091 by Dominicus and Munnio, and the "banner" display texts cited by Walker were part of the 1091 Beatus commission directed by Munnio. But its figures were added only in 1109 by Petrus. Munnio did not do figures. If Silos is going to be a focus, Munnio is an important figure. Walker recognized him: toward the end (127) of her analysis of the illumination of Add. MS 30848 she mentions in passing that Boylan attributed this manuscript to Munnio, but without indicating whether she endorsed that conclusion. The Munnio connection deserves more attention. Not only does it link one of her manuscripts with a firm date, but the 1091 Beatus contributes to the history of the Silos scriptorium: it is replete with colophons making sure that the Silos scriptorium is properly recognized. Munnio's involvement touches an even larger point, that of the kind of direction behind the change of liturgy. Walker tends to favor, without actually spelling out its operation, the idea of some kind of central control, as we saw in her suggestion of a verbal directive for the inclusion of man-dragon initials, her rejection of a Moissac scribe for the Facundus Missal in favor of a Cluniac origin, and also in her suggestion that use of the Romanesque style "may have been a requirement when decorating the central text of the new liturgy." (218) The fact that Add. MS 30848 with Munnio type "banner" headings is the very manuscript that includes, uniquely, readings from the Beatus Commentary suggests that Munnio, a monk, was responsible for not only the illumination of the Commentary, but at least part of its textual content as well.

Since she cites Meyer Schapiro and is inclined to want to identify her manuscripts with the monastery of Silos, Walker's conclusion that there "is scant evidence for enthusiastic acceptance of the new liturgy among those who produced the new manuscripts for many years after the initial edict" inevitably calls to mind Schapiro's famous thesis that the changes at Silos, most particularly the liturgical change, were resisted ("From Mozarabic to Romanesque at Silos" in his Romanesque Art). She also shares with Schapiro a debatably high estimation of the abbey's importance in the eleventh century: "The abbacy of Domingo at Silos (1041-1073) was a time of great development, and of powerful political connection. It seems improbable that such a major institution as the reformed abbey would not have had its own scriptorium. Silos was sufficiently important--and sufficiently wealthy--to have Domingo's body translated in the presence of Alfonso VI in 1076'" (119) The Vita of Domingo, written not long after his death, does not, in fact, claim that Alfonso VI was present for the translation, but it uses the exaggerations of eminence customary in hagiographic literature. The documents of the abbey do not support the exalted position for it assumed by Schapiro. Schapiro's evidence for resistance to the changes taking place under Alfonso VI was the style of the Beatus Commentary of 1091-1109. He argued that the outdated, Mozarabic, style had been resurrected to signal homage to the traditional Hispanic church on the part of some monks. Walker's conclusion from the appearance of her liturgical manuscripts that the change was not popular comes across as a variation on this problematic approach. Her conjecture that, "Even if the abbot and other leading figures agreed to implement the change, that does not mean that they or the monks, who were also scribes and artists, were necessarily in favour of it" (220) also seems to echo Schapiro, but even he accepted the fact that "it is probable that Silos played an active part in the change of religious rule in Castile." (op. cit., 61) San Millan de la Cogolla produced, as she duly notes, two texts defending the ancient liturgy. But must we assume that monks in Silos and elsewhere were against the change? In any case, the style or quality of manuscript illumination seems unreliable evidence.