Revisiting the Umbrian retro-choir

Donal Cooper

doi:10.4000/books.efr.21290

Revisiting the Umbrian retro-choir

Plurality and choice in the medieval Franciscan church interior

Donal Cooper

p. 81-150

Résumé

Recent research has clarified many aspects of the pre-Tridentine mendicant church interior, notably the importance of tramezzo rood screens, the elevated display of large-scale painted panels and – the focus of the present book – the location of choir precincts. It has, however, also revealed the complexity and diversity of liturgical furniture and spatial arrangements. This degree of flexibility cautions against reading Renaissance church interiors in crudely functional terms. The introduction of apsidal choir precincts or so-called “retro-choirs”, once regarded as a fundamental characteristic of Tridentine rearrangements, is now understood as an extended process responding to a number of distinct factors. This paper examines some of the earliest evidence for mendicant retro-choirs associated with Franciscan churches in Umbria, stretching back into the thirteenth century. I first highlighted this anomalous phenomenon in an article published in 2001 in the Journal of the Warburg and Courtauld Institutes: I can now introduce additional archival material to both consolidate and nuance those findings. The ambiguous case of Sansepolcro – in the modern region of Tuscany but falling within the medieval Franciscan province of Umbria – suggests that the adoption of the retro-choir respected the Order’s provincial boundaries. Despite sharing an identical liturgical framework, friars in the Tuscan and Umbrian provinces developed very different modes for arranging their church interiors. These differences are reflected in both the architecture of the churches and the artworks commissioned for them. This paper reassesses the evidence for provincialism in Franciscan church configuration in central Italy, asks how this was possible, and considers what such distinctions might have signified for a contemporary audience.

Entrées d’index

Keywords : Umbria, Franciscan Order, altarpiece, retro-choir

Note de l’auteur

I am very grateful to Joanne Allen, James Banker, Machtelt Brüggen Israëls, Joanna Cannon, Sible de Blaauw, Silvia De Luca, Andrea De Marchi, Alessandro Delpriori, Lorenzo Donati, Dillian Gordon, Michael Gromotka, Giacomo Guazzini, Elvio Lunghi, Paola Mercurelli, Francesco Piagnani, Alberto Maria Sartore, Maria Elena Sigismondi, Roberto Sigismondi, Rita Silvestrelli, Gail Solberg and Lucilla Vignoli for reading drafts of this essay, offering expert advice and comments, and providing generous assistance with illustrations and bibliography. Antonella Pinna of the Servizio Musei, Archivi e Biblioteche of the Regione Umbria kindly helped to source relevant groundplans. I am also indebted to Haude Morvan for the invitation to revisit this material and her perceptive editing.

Texte intégral

Fototeca Servizio Musei, archivi e biblioteche della Regione Umbria

1When it came to designing and furnishing their churches, the Franciscans in Umbria seem to have done things differently – differently, indeed, from every other province in their Order. From an early date, probably even as their first substantial church buildings began to rise around the middle of the thirteenth century, the Friars Minor of the Order’s Umbrian province – the Provincia Sancti Francisci – sited their conventual choir precincts behind high altars (fig. 1), not in front as was common practice elsewhere. Crudely summarised, this was the argument advanced in my 2001 article Franciscan Choir Enclosures and the Function of Double-Sided Altarpieces in pre-Tridentine Umbria, published in the Journal of the Warburg and Courtauld Institutes1. The Umbrian evidence I presented there pushed the origins of the so-called retro-choir – traditionally associated with the comprehensive renewal of ecclesiastical fabric during the Counter Reformation or the more sporadic experiments of Quattrocento church architecture – back to the late medieval period, potentially over three centuries before the Council of Trent. The Umbrian approach to sacred space also informed the development of a distinctive provincial tradition of high altarpiece design, with double-sided panels (low retables at first, followed by lofty polyptychs) intended to serve both a lay public in the nave and a conventual audience in an apsidal choir.

2The occasion of a conference specifically dedicated to the configuration of choirs in mendicant churches offers an opportunity to revisit the precocious phenomenon of the retro-choir in Franciscan Umbria. A wealth of new research over the two decades since my article appeared has, if anything, underlined the anomalous status of the Umbrian case within the broader landscape of late medieval church fabric and liturgical configurations. At the same time, fresh evidence at a regional level suggests a more complex and varied picture, revealing exceptions to the retro-choir model within the Order and early appropriations without, as the phenomenon quickly escaped the confines of the Franciscan movement. This chapter will reassess my original findings to see if they can be sustained or refined, integrating relevant research undertaken in the intervening years, including significant new archival discoveries by myself and by other scholars.

3The point of departure for my 2001 essay was the need to reconcile a glaring contradiction in the scholarship to that date between the literature on double-sided altarpieces (which for some time had simply assumed the existence of retro-choirs in medieval buildings2) and the work of Renaissance architectural historians (who characterised the phenomenon of the apsidal choir as an innovation of the Quattrocento and early Cinquecento attributable to leading architects like Brunelleschi, Michelozzo and Giuliano da Sangallo, alongside their patrons)3. The article’s conclusion advanced some ambitious assertions for the Umbrian context, notably that with regard to choir arrangements “sixteenth-century Counter Reformers were simply borrowing solutions that had already been perfected in medieval Gothic churches”, and that “the arrangement of almost every conventual church in Italy would ultimately bear [the] imprimatur [of the Umbrian model]”4. Due in part, perhaps, to these bold claims, my article quickly attracted scholarly attention, generating lively debate at the conference on Lo Spazio e il culto held at the Kunsthistorisches Institut in Florence in March 20035. Surveys of Italian liturgical space by Michele Bacci (2005) and Marcia Hall (2006) subsequently flagged Franciscan Umbria as the only notable exception to the pre-Tridentine norm of nave choirs6.

4While more examples of apsidal choirs and the clearance of nave precincts have been identified from the Quattrocento, the Umbrian material still constitutes the most significant regional anomaly and remains by far the earliest7. Indeed, the Umbrian province looks more, not less, unusual in the light of extensive research undertaken on Italian liturgical arrangements and ecclesiastical furniture over the past two decades8. The increased attention paid to tramezzo screens and nave choirs as key determinants for visibility and access within church interiors means that their presence can be confidently asserted – at least for mendicant and monastic churches – across most of the other Italian regions9. They constituted the dominant mode for organizing Italian sacred building throughout the late medieval period, and even into the Quattro- and Cinquecento, notwithstanding the gradual spread of retro-choirs from the mid-fifteenth century on. The norm is exemplified in its basic guise by the so-called “Fra Giovanni da Pistoia” parchment plan for San Francesco, Arezzo (fig. 2), which depicts that church in the fourteenth century replete with its liturgical furniture10. The Arezzo plan gives us a good sense of what we would expect to find inside a Tuscan church, with a masonry tramezzo screen splitting the building into unequal halves, a choir precinct in the upper nave and the high altar in the cappella maggiore beyond equipped for celebration ad orientem11. The template could be elaborated for larger churches with aisles or by the addition of arcades to the screen wall to support preaching platforms and elevated altars. But the same fundamental logic – a hierarchical sequence of spaces culminating with the high altar sited in the main apse – seems to have held good for conventual buildings across the peninsula, except for Umbria, until the mid-1400s.

Photo: Archivio storico della diocesi di Arezzo

5This chapter returns to the Umbrian context and expands the pool of Franciscan examples to cover centres omitted from my 2001 essay, seeking a more representative sample of the 74 convents that constituted the Provincia Sancti Francisci by the early fourteenth century (fig. 3)12. New evidence concerning choir precincts is assembled in the opening two sections: the first section presents fresh material that broadly affirms the retro-choir hypothesis, the second introduces findings that complicate and nuance it. The third section considers the production of high altarpieces for the Provincia Sancti Francisci, extending the scope of my 2001 treatment to include single- as well as double-sided paintings together with examples stretching into the early Cinquecento. The fourth and final section draws comparisons with the practices of rival religious orders and the secular clergy in the region to explore how liturgical practice may have seeped across institutional divides to form a more general, regional trend, as opposed to one associated particularly with the Franciscans. Reviewing the current state of research, the conclusion anticipates future avenues of enquiry and highlights the complexities and variations presented by the evidence, subtleties that need to be acknowledged in any effort to draw patterns of practice. It also considers the meanings and motivations that may have informed particular choices. The “papal” arrangement of the Upper Church at Assisi, complete by 1253 if not somewhat earlier, remains an obvious candidate for the origins of the Umbrian retro-choir (fig. 4), but we still have little idea regarding the connotations or symbolisms that attached to different configurations of liturgical space13.

Groundplan by author adapted from Giuseppe Rocchi

Revisiting the evidence for the early retro-choir

6The best place to begin our analysis of Franciscan Umbria is San Francesco in Sansepolcro (ironically the only church in the Order’s province that sits across the modern regional border inside Tuscany). Thanks to the Sassetta project led by Machtelt Brüggen Israëls and completed in 2009 we now know more about this church than any other foundation in the Provincia Sancti Francisci apart from Assisi (fig. 5)14. Notarial documents found by James Banker and myself over the course of that project have confirmed the location of the conventual choir in the cappella maggiore in more unambiguous terms than I could have hoped for when writing my JWCI article15. In 1352 an earthquake had felled the main chapel vault, badly damaging the choir below – an indication that the stalls were already located behind the high altar before this date. At first the friars concentrated their resources on repairing the vault but in 1376 they began to release money for a new set of choir stalls and in 1378 their Operai signed a contract with the local stonemason Luca di Bernardo to reficere et rehedificare de novo chorum dicte ecclesie post altare maius ecclesie supradicte16. A later contract agreed in 1392 with the woodcarvers Giovanni and Angelo Schiatti for ongoing work on the stalls also specified a “sedia grande” at the choir’s centre, presumably for the convent’s guardian17. Assuming that this more eminent chair was set in the middle of the bank of stalls at the back of the choir, directly below the apse’s main double-lancet window, the choir’s layout has ramifications for our understanding of Sassetta’s later polyptych over the high altar. Following the installation of the Sienese artist’s altarpiece in May 1444 – a commission to which we will return below – the convent’s guardian would have found himself aligned, and by extension associated with the radiant representation of St Francis at the centre of the back of Sassetta’s great double-sided structure (fig. 30).

7Our reconstruction of the Sansepolcro church assumed the presence of a nave tramezzo screen; references in the 1583 Apostolic Visitation for the town and earlier notarial acts were not unequivocal on this point, but suggested enough for the screen to be included18. This reminds us – as Sible de Blaauw has rightly observed – that the development of the retro-choir and the clearing of the nave space were distinct processes before the Renaissance, even if their effects may appear with hindsight to converge19. Another aspect of the Franciscan patronage of liturgical furniture that the Sassetta project revealed was the remarkable patience and persistence displayed by the local friars when it came to re-equipping their church. The 26-year hiatus between the 1352 earthquake and the initial contract for a new choir in 1378 may be understandable given the severe structural damage sustained by the church, but work on the stalls dragged on until just before 140020. The friars then waited another quarter century until 1426 to initiate a project for a new high altarpiece which only arrived in 144421.

8Sansepolcro is an unusual and revealing example because its interior followed the Umbrian liturgical template despite its architectural superstructure being in certain key respects more Tuscan, for example in the rectangular cappella maggiore (although the chapel’s notable size relative to the nave and its distinctive fenestration suggest that it was designed from the outset to accommodate a retro-choir)22. As rebuilt in the late Trecento, the ribs of the cappella maggiore vault were sprung from corbels, which would have allowed choir stalls to have been placed flush against the walls below23. Elsewhere across the Umbrian province, retro-choirs were associated with polygonal apses ultimately inspired by the Upper Church at Assisi24. In two of the more remote convents – Gualdo Tadino (fig. 1, 6, 7 and 8) and Cascia (fig. 9 and 10) – early sets of stalls still survive in situ to provide an idea of how choirs and apses could be integrated together25. Here too, the vault shafts were corbelled, thereby leaving blank wall faces free of engaged columns for the positioning of choir stalls below. Thanks to new archival discoveries, a retro-choir can be confirmed for the central polygonal apse in San Francesco, Città di Castello (fig. 11), for which few documents were available in 2001. Although not as rich as the peerless archives for nearby Sansepolcro, the Città di Castello notarile has been yielding a steady stream of atti relating to the local Franciscan church and its furnishing. Documents from 1449 and 1454 for the renewal of two of the three windows in the main apse make it clear that these apertures were directly above the friars’ choir stalls which were in turn located behind the high altar (a parte posteriori altaris maioris)26. These sources confirm that the church possessed a retro-choir by that date and had probably done so for some time, as in the 1480s these stalls were old enough to be replaced by a new set27. Other Quattrocento notarial acts in the Città di Castello archives refer to a pergulum or pulpitum (probably but not necessarily one and the same structure28) and indicate that the nave space could be divided by gender – suggestive but not conclusive evidence for the existence of a tramezzo screen29.

9Surviving mural decorations provide further evidence for retro-choirs in Umbrian churches. Benozzo Gozzoli’s frescoes in the main apse of San Francesco, Montefalco, completed in 1452, are a case in point (fig. 12). The busts of illustrious Franciscans that populate the frieze along the bottom of the narrative scenes were surely intended to oversee the convent’s latter-day friars in their stalls below30. Documents published by Silvestro Nessi in 2002 reveal testamentary bequests for a new set of choir stalls in the church from the years 1455-5731. While their location is not specified, the timing hard on the heels of Gozzoli’s frescoes suggests a link with the renovation of the cappella maggiore. The vault shafts rest on corbels directly above the present (nineteenth-century) stalls. The church was in use by the early 1340s and the conservation of Gozzoli’s cycle following the 1997 earthquake revealed an earlier level of fourteenth-century frescoes below the Quattrocento murals32. Their full extent is unknown and it is unclear whether they originally descended to the floor of the apse, but they are unlikely to have left sufficient room for banks of choir stalls33. Silvestro Nessi’s archival work and Alessandro Delpriori’s analysis of the church’s interior now provide convincing evidence for the early existence of a tramezzo screen dividing the nave (fig. 13), on which the large crucifix attributed to the Maestro Espressionista di Santa Chiara was mounted34.

10The architecturally sophisticated church of San Fortunato, Todi, begun in 1292, possesses a seven-segment apse free of surviving frescoes but with beautifully corbelled vault shafts (fig. 14 and 15). The apse now hosts choir stalls constructed between 1581 and 1590 by Antonio Maffei da Gubbio35. Earlier documents for repairs to the church’s fenestration describe windows above the choir, a further indication of an apsidal precinct36. San Fortunato was unusual amongst the Order’s Umbrian foundations in that its high altar doubled as the shrine for a long-established cluster of civic cults – the local saints Fortunatus, Calixtus, Cassianus, Romana and Degna – rather than a recent Franciscan beato. The removal of their relics during the renovation of the high altar in 1580 attracted the attention of local antiquarians who in turn provide indirect evidence for the former, medieval arrangement of the interior37. Writing in 1597, Giovambattista Possevino recalled that the relics of Todi’s three patron saints had been removed from the high altar, which had formerly occupied “a large part of the church”38. Giovan Battista Guazzaroni (d. 1624) was more precise, specifying that the reconfiguration in 1580 was intended to move the high altar closer to the choir, indicating that the stalls were already located in the apse prior to this date39. Possevino implied that the arrangement dismantalled in 1580 dated back to 1301, strongly suggesting that it was integral to the architectural plan begun in 129240.

11The decision to move the high altar back towards the apse runs contrary to the general trend of sixteenth-century reconfigurations of Italian church interiors, where high altars were as a rule moved forwards to create extra space in their wake for new apsidal choir enclosures that had not existed before but were now needed following the clearance of nave precincts and screens41. As at Todi, the high altar of the main Franciscan church in Perugia, San Francesco al Prato, was moved back towards the church’s polygonal apse in the sixteenth century, and it was clarification of this rearrangement that underpinned my 2001 reconstruction of that church’s interior with a retro-choir (fig. 16)42. The re-consecration of San Francesco al Prato’s high altar in 1536 seems to have marked the culmination of a significant renewal of the church’s liturgical fabric43. A list of floor tombs in the church composed shortly before in 1532 recorded a row of new burials laid out where the high altar had formerly been (“dove prima era lo altare grande”), presumably referring to the medieval altar sacrificed for the new altar block that would be consecrated four years later44. Careful analysis of the church’s topography and related sepoltuarii placed these tombs close to the juncture of the crossing bay and nave, whereas the 1536 high altar is marked on an early eighteenth-century plan at the mouth of the apse45. The rearrangement of San Francesco al Prato thereby moved the high altar of the church back towards the apse by as much as 10 metres. Thanks to a contract discovered by Alberto Maria Sartore, we also know that this reconfiguration encompassed a new set of choir stalls, which the friars commissioned in 1530 with an expected completion date in 153446.

12The pre-1530 arrangement at Perugia would have allowed for a more spacious choir arrangement behind the high altar and a number of documentary notices suggest that the previous conventual choir was indeed located in the apse. In 1468 the church’s apse (or “tribuna”) was associated with its choir in a payment for shoring up the building’s perennially problematic foundations and in 1476 repairs were made to the windows of the choir (fenestra vitrea chori ipsius ecclesie)47. The church’s architecture further supports the retro-choir hypothesis. Pope Innocent IV laid the foundation stone during his Perugian residence of 1251-53 and the design was clearly modelled on the newly completed Upper Church at Assisi, which Innocent consecrated in May 125348. San Francesco al Prato’s vaulted Latin cross groundplan and occidented site both acknowledge the Order’s Mother Church at Assisi. The Perugian architects, however, modified the template of the Upper Church by corbelling their vaults and adopting a seven-segment as opposed to a five-segment apse, a departure that facilitated a significantly deeper apsidal space (circa 10 metres as opposed to circa 7 metres). This innovation seems to have appeared first in Perugia and was later copied at San Fortunato, Todi, where the apse is circa 12.5 metres deep49.

13The foregoing sweep of recent research confirms the historical reality of the retro-choir in medieval Umbria. At Sansepolcro the new archival evidence is unequivocal on the location of a conventual choir behind the high altar by the middle of the fourteenth century. For nearby Città di Castello notarial acts document a similar arrangement by the middle of the fifteenth century. These examples can be added to the surviving Trecento and early Quattrocento stalls in the apses at Gualdo and Cascia. Furthermore, the architectural evidence of Franciscan churches suggests that the practice of retro-choirs predates the available archival sources and had its origins in the mid thirteenth century. It is tempting to think that the Umbrian template of large and deep apsidal vaults with corbelled shafts that was established by San Francesco al Prato in the 1250s was designed to accommodate banks of choir stalls behind high altars.

Ambiguities and exceptions

14The picture for Umbria is not, however, altogether uniform. While the presence of retro-choirs in Cascia, Città di Castello, Gualdo Tadino, Montefalco (at least in the fifteenth century), Perugia, Sansepolcro, and Todi can be securely asserted on the basis of surviving fabric or archival evidence, new discoveries elsewhere in the Provincia Sancti Francisci point to different arrangements and the existence of nave choirs similar to those found in other Franciscan provinces. The Order’s church in Gubbio is one such case. San Francesco, Gubbio, was one of the earliest large-scale Franciscan churches to be built in Umbria after the Basilica at Assisi50. It differs markedly from other foundations in the Provincia Sancti Francisci in its three-aisled plan. Its central apse is also small and shallow in relation to the overall scale of the building compared to the other Franciscan churches discussed above. Floor repairs in March 1983 uncovered the foundations of a dividing wall in the middle of the nave which have been identified as the remnants of a choir barrier (fig. 17), although, as we have seen above, there are other cases where retro-choir arrangements might have been combined with nave tramezzi51. It may be significant that the cappella maggiore at Gubbio was not associated with choir stalls in the wording of a testamentary bequest for window repairs in 1420, unlike the similar documents for Città di Castello and Perugia cited above52. While the pre-Tridentine documentation for San Francesco, Gubbio is disappointingly thin, Debora Lilli concluded from her detailed investigation of the church’s later baroque remodelling that a nave choir had probably been removed some time before the extensive renovations undertaken in the seventeenth century53.

15Alessandro Delpriori has recently expressed doubts regarding the feasibility of a retro-choir in San Francesco, Trevi (fig. 18). The church at Trevi would be a more surprising exception than Gubbio, as its plan is more characteristically Umbrian, with a single nave terminating in a central polygonal apse flanked by shallow rectangular chapels (comparable to San Francesco at Montefalco)54. At Trevi the five-segment vault of the cappella maggiore is sprung directly from corbelled feet with no descending shafts. As Delpriori has observed, however, the recent conservation of the main apse suggests that its Trecento fresco decoration extended down to floor level (fig. 19), rendering the early installation of an apsidal bank of stalls unlikely55. Further clues can be gleaned from the available notarial documentation. In 1433 a testamentary bequest left a green cloth dicte ecclesie Sancti Francisci pro apparatu altaris, ubi est crucifixus, a parte inferiori ipsius ecclesie, which would suggest that the large crucifix today located in the apse was originally over a dividing screen across the middle of the nave56. In 1480 a contract for a range of structural interventions in the sacristy, refectory and dormitory of the adjacent convent mentioned a stair and parapet that led from the choir to the cloister57. The present cloister, much rebuilt in the seventeenth century, is located to the north of the nave, but does not extend as far as the apse58. The local chronicler Ser Francesco Mugnoni recorded the audience given by the Order’s Minister General Egidio Delfini d’Amelia in 1502 from a “sedia” in San Francesco located between the entrance to the church (presumably the large portal located midway down the southern nave wall) and the high altar, although whether this was part of a larger choir precinct is unclear from Mugnoni’s account59. The fragmentary documentary evidence presently available for San Francesco, Trevi is not conclusive, but does not contradict Delpriori’s argument for a nave choir on the basis of the apse frescoes.

16Elsewhere in Umbria, the architectural evidence of a number of Franciscan churches suggests the presence of retro-choirs. In San Francesco, San Gemini, a spacious five-segment apse with corbelled vault shafts would have been well-suited to housing choir stalls against its lower walls (fig. 20)60. A similar situation is found in San Francesco, Narni, and here the five-segment apse is so broad and deep in relation to the main body of the church (in this case, unusually, a three-aisled nave divided basilica-style by colonnades) that it is difficult to explain its scale without accepting a retro-choir hypothesis (fig. 21, 22 and 23)61. In San Francesco, Montone, an elegant five-segment apse rests on richly carved corbels. The church was built in the early 1300s and the apse is now filled with a bank of thirteen stalls datable to the end of the fifteenth century (fig. 24 and 25)62. However, the earlier fresco decoration of the Montone apse by the Ferrarese painter Antonio Alberti, datable by heraldic evidence to 1422-24, incorporated a fictive curtain that extended to the base of the apse, behind the present stalls63. Furthermore, these drapes were painted over an even older fresco layer that depicted one of the original cross motifs dating from the consecration of church around the middle of the fourteenth century, so it seems unlikely that the Montone apse was built to accommodate a retrochoir at the outset despite its conveniently corbelled shafts (fig. 26)64.

17Archival acts discovered by Stefano Felicetti and published in 2001 document commissions for the painting of high altar chapels in the Franciscan churches in Amelia (1371 and 1431)65, Spello (1434)66, and Bevagna (1498-99)67. In these cases the cappella maggiore is referred to as the tribuna, the cappella magna, the cappella principalis or the chapel of the high altar, but is not related to a choir space (unlike the archival notices examined above for Città di Castello and Perugia). It is risky to place too much weight on the usage of architectural terms in notarial documents, where their meaning could be more or less fluid and determined by legal rather than descriptive considerations. Further archival research on these centres may yield firmer data. The churches at Amelia (fig. 27) and Bevagna were heavily remodelled in the eighteenth century with entirely new baroque interiors68. Both had originally been built with five-segment apses whose inner faces are now masked by later vaults. Any traces of the documented fresco schemes are today buried within the thickness of the apsidal walls and it is impossible to tell whether their vault shafts were corbelled or not. The Spello apse survives, but any surviving fifteenth-century frescoes are today covered by the pseudo-medieval murals executed in 191469. The balance of evidence is reversed for the remote church of Santa Maria in Vallo di Nera. Here there are few available documents but the medieval mural decoration is exceptionally well preserved (fig. 28)70. The single nave church terminates in a simple square apse which contains an extensive fresco decoration executed by Cola di Pietro from Camerino and Francesco di Antonio from Ancona dated by their inscription of co-authorship to 1383. The scheme includes fictive panelling below the narrative scenes that extends down to the floor of the cappella maggiore.

18We thus have a number of churches in the Provincia Sancti Francisci – Montefalco, Montone, Trevi, and Vallo di Nera – where frescoes on the lower walls of the main apsidal chapels render the presence of retro-choirs in the fourteenth century unlikely. At Montefalco, we can be fairly certain that the church was reconfigured with a retro-choir in the 1450s and the same may have been the case with Montone around 1500, but neither church was built to accommodate apsidal choirs in the first instance. The apsidal chapels at Montefalco, Montone, Trevi and Vallo di Nera were all built with corbelled vault shafts, suggesting that the popularity of this architectural feature was not necessarily linked to the phenomenon of retro-choirs. This needs to be borne in mind for other Franciscan churches – like those at Narni and San Gemini – where archival sources are so far lacking and polygonal apses with corbelled shafts constitute the main evidence for retro-choir arrangements. In this light, it is telling that the vault shafts of the main apse in San Francesco, Gubbio (alongside Perugia one of the oldest of the Provincia’s major churches after Assisi itself) are not corbelled but those in the side apses are. In other cases – Amelia, Bevagna and Spello – the medieval apses no longer survive and we have only documents to work with. None of these archival sources associate the high altar chapels concerned with conventual choir stalls, although the description of the 1498 Mezzastris commission at Bevagna seems to reprise Gozzoli’s cycle at Montefalco down to the bust-length gallery of Franciscan luminaries. The evidence is cleary mixed, but points to a significant number of churches in the Provincia Sancti Francisci being built with nave choirs similar to those in other provinces during the thirteenth and fourteenth centuries. Evidently the Umbrian Franciscans could choose between two (at least) modes of organizing their churches.

Double- and single-sided high altarpieces in Franciscan Umbria

19In my JWCI article I interpreted the series of double-sided altarpieces commissioned for the high altars of Franciscan churches in Umbria as a distinctive symptom of the regional pattern of retro-choirs, developing the thesis first advanced for Perugia by Henk Van Os in 197471. Here too, a survey of subsequent research reveals a more mixed picture. On one hand, a number of combined retro-choir / double-sided altarpiece arrangements can be confirmed with greater confidence. At the same time, other churches in the Provincia Sancti Francisci certainly had single-sided high altarpieces, and it is not always clear what kind of choir was configured around them.

20The Sassetta project clarified in exacting detail the double-sided carpentry of the great high altarpiece that the Sienese master completed for the church of San Francesco, Sansepolco in 1444 (fig. 29 and 30). To judge from the documentary record, Sassetta’s altarpiece was set towards the rear of the deep altar mensa that still survives in Sansepolcro72. The “front” of the polyptych could therefore be regarded as the backdrop for the Mass, while the rear would have faced the friars gathered in their choir stalls. As noted above, the existence of a “sedia grande” at the centre of the choir precinct very likely aligned the convent’s guardian with the dominant figure of Saint Francis on the back of the altarpiece73. If there was no provision for Mass to be celebrated to the rear of the polyptych facing the choir, how did the friars participate in the liturgy? The structure of the altarpiece’s predella may hold the key, as the central panels from both faces of the polyptych’s base do not survive and the comprehensive iconographic programme agreed in 1439 made no provision for their imagery74. One possibility is that the mid-point of the predella box integrated a Eucharistic tabernacle for the reservation of the Host, or indeed a shuttered opening that may have allowed the Host to be seen from the rear of the polyptych75. As we shall see, other examples offer circumstantial support for both these options. Sassetta’s altarpiece was finally supplanted by a Tridentine-style Eucharistic tabernacle in 157876.

Reconstruction: Machtelt Israëls et al., three-dimensional drawing by Andrea Santacesaria, Opificio delle Pietre Dure, Florence, rendering by Giacomo Tenti, Culturanuova, Arezzo (Villa I Tatti, The Harvard University Center for Italian Renaissance Studies)

21For nearby Città di Castello I have assembled further evidence for the existence of a double-sided polyptych in the local Franciscan church, painted by Spinello Aretino (perhaps in two phases) in the years to either side of 140077. To judge from the three preserved fragments and the surviving high altar mensa, this would have been even larger than Sassetta’s later polyptych at Sansepolcro78.

22The third in this series of major double-sided polyptychs was painted by the Sienese artist Taddeo di Bartolo and installed over the high altar of San Francesco al Prato, Perugia in 1403. Building on her earlier publications, Gail Solberg has recently clarified the structure of Taddeo’s altarpiece and identified the final saint from the main register (fig. 31)79. It is generally accepted that the 1403 polyptych replaced a much earlier retable painted by the anonymous Maestro di San Francesco around 1272 (the date inscribed on a large crucifix painted by the artist for the same church). Recent publications have confirmed Dillian Gordon’s reconstruction of the preserved panels of the Maestro di San Francesco’s altarpiece as fragments of a low, double-sided retable (fig. 32 and 33)80. Only one wing of the dossal survives but its early date, within 20 years of the church’s foundation, is a further argument for San Francesco al Prato having possessed a retro-choir from the outset.

Reconstruction: D. Cooper after Dillian Gordon

23Like Sassetta’s altarpiece at Sansepolcro, the central section of the predella from Taddeo di Bartolo’s polyptych does not seem to survive, leaving open the possibility that it may have incorporated a shutter or opening81. Taddeo’s altarpiece was clearly an encumbrance when San Francesco al Prato was required to host the papal liturgy during Julius II’s passage through Perugia in 1506. The state of the Franciscan church did not impress the papal master of ceremonies, Paris de Grassis, who found that the choir was too old and fragile to comfortably seat the cardinals or to support the platform for the papal throne. Moreover, the 1403 polyptych could not be removed due to the protests of the local Franciscans: Paris, clearly irritated, added that the friars were exceptionally unhelpful (fratrum maxima malignatas fuit)82. Flanked by ground-breaking altarpieces by Raphael, Taddeo’s painting must have appeared dated to early Cinquecento viewers and a 1533 testamentary legacy of 25 florins expendendos in ornamentis tabule altaris maioris et principalis dicte ecclesie may signal a desire to either update or replace the medieval polyptych83. Eventually in 1592 the friars commissioned an elaborate Eucharistic tabernacle from the local painter Fiorenzo di Giuliano. In the contract they asked Fiorenzo to make the base of the tabernacle high enough so that the friars could see the sacrament from the choir84.

24A succession of double-sided altarpieces can also be reconstructed for the second Franciscan church in Perugia, San Francesco al Monte set on the hill of Monteripido outside the city walls. Dillian Gordon convincingly traced a double-sided retable now in the Galleria Nazionale dell’Umbria by the Maestro di Paciano and datable to c. 1322 back to Monteripido (fig. 34 and 35)85. This altarpiece has evidently lost a lower register or pseudo-predella; the truncated tops of an enthroned Christ and two adoring angels are still visible below the central Virgin and Child on the “front” of the retable. By the middle of the Quattrocento the friars at Monteripido and their benefactors were considering a replacement. In his 1452 testament, Tommaso di Paolo dei Ranieri left money for a new panel in coro et super altare maioris. Tommaso allowed the friars to either paint both sides of the tabula or only the face ubi dicitur missa, confirming that the panel was potentially double-sided and also that one side was designated for the celebration of the Mass while the other was implicitly not, a similar situation to that suggested above for Sansepolcro86. It is unclear whether Tommaso’s legacy ever led to a commission but in 1502 the friars contracted Pietro Perugino to paint a panel that was evidently already in place over the high altar, described in the contract as the tabulam existentem super altare magnum dictae ecclesiae87. Perugino was asked to paint a Coronation of the Virgin (fig. 36) on the side “verso la chiesa delle donne” which in a male conventual church like San Francesco al Monte can only refer to the lower nave88. The other side, presumably facing the choir, already bore a large carved-wooden crucifix for which Perugino painted accompanying figures of the Virgin, Saint John, the Magdalen and Saint Francis (fig. 37)89.

Photo: Soprintendenza Archeologia, belle arti e paesaggio dell’Umbria

25At Gualdo Tadino the local Franciscans commissioned a single-sided altarpiece (fig. 38) from the Foligno artist Niccolò Alunno in 1470 for their church of San Francesco, which – as we have seen above – still conserves its Trecento apsidal choir90. It is unclear why the Gualdo friars chose a single-sided polyptych over a double-sided one; cost may have been a consideration91. However, they did ask Alunno to include an open arcade between the predella and the main register which would have allowed a partial view of the liturgy on the altar from the vantage of the apsidal choir92. This arcade was later removed, probably in the eighteenth century when the altarpiece was finally shifted from the high altar to a side chapel. The restoration carried out in 1928 sought to reintegrate this feature copying the format used by Alunno in 1483 for his high altarpiece for the cathedral at Nocera Umbra93. This intervention was reversed in the 1950s but is recorded in early photographs. At Gualdo the predella still preserves a central tabernacle for the Host with adoring angels to either side, although whether this could be accessed from behind is unclear94.

26Other single-sided high altarpieces can be associated with Franciscan churches in Umbria. Perhaps the earliest is the dispersed polytpych attributed to the Maestro di Figline (fig. 39), the major elements of which can be identified in the Ranghiasci collection in Gubbio in the 1870s95. Measuring at least two and half metres wide, the altarpiece must have been commissioned either for a high altar or for a significant conventual or chapel space96. The Gubbio provenance, the prominence of Saint Francis in the main register and the fact that the Maestro di Figline seems to have worked almost exclusively for the Order have led scholars to assume a Franciscan commission from Umbria97. Miklós Boskovits and Francesca Mambelli have suggested that the altarpiece was painted for a church in Gubbio or nearby, but the widely-agreed dating of the surviving panels on stylistic grounds to the 1320s is hard to reconcile with the rather later testamentary bequest from 1345 of 100 soldi to the friars of San Francesco in Gubbio pro adiutorio causa fieri faciendi unam tabulam pictam causa ponendum super altare maiori dicte ecclesie (a project about which nothing more is known)98.

27The other full-length saints from the Maestro di Figline’s altarpiece may offer a clue to its original provenance. Saint Peter serves as a common pendant for Saint Francis, but the apostles Philip and James, commemorated on a shared feast day on 3 May, are unusual in a Franciscan context. Their inclusion on the altarpiece’s main register may signal a connection with the dedication of a church or chapel, and two Franciscan foundations in Umbria can be suggested on this basis. The Order’s church in Amelia was initially dedicated to Saints Philip and James before assuming the title of San Francesco over the course of the fourteenth century99. In 1324 the seal of the Amelia convent was described as bearing the images of the two apostles100. A separate case can probably be made for the Order’s extramural foundation of Santi Filippo e Giacomo at Montefalco, which the friars abandoned in 1336 in favour of a more secure site within the town walls101. Despite its curtailed history, Santi Filippo e Giacomo was evidently a sizeable church and important enough to generate the commission of the large painted cross by the Maestro Espressionista di Santa Chiara, an artist whose principal Umbrian base – like the Maestro di Figline – seems to have been at Assisi102. The Montefalco convent was particularly well-connected with Assisi in the early years of the Trecento thanks to Fra Francesco Damiani, a local Montefalco friar who served as Umbrian Provincial Minister in the second decade and whose elder sister, Chiara, was a candidate for canonization103. A third, more remote, possibility is that the Maestro di Figline’s “Franciscan polyptych” was not a Franciscan project at all104. Current reconstructions have assumed that the Maestro di Figline’s polyptych was single-sided although the condition of the surviving panels has led some scholars to suspect that the altarpiece may originally have been double-sided105. Detailed analysis of the surviving panels would doubtless resolve many aspects of the original carpentary, but other single-sided high altarpieces were certainly commissioned for the Provincia Sancti Francisci.

28Another early and single-sided altarpiece can be linked with Montefalco, this time with the new church of San Francesco constructed within the town walls following the demolition of Santi Filippo e Giacomo. This pentaptych with scenes from Christ’s Passion attributed to the Maestro di Fossa is now in the Papal Apartments in the Vatican but was traced back to the Montefalco church by Silvestro Nessi and has been securely dated to 1336 thanks to a recently discovered inscription106. It is tempting to connect the Vatican retable with a solitary Franciscan scene by the same artist, formerly in the collection of Osvald Sirén and now known only through an auction house photograph107. Both Alessandro Delpriori and I believed that the Sirén scene could have been an orphaned remnant of a cycle that was originally on the “back” of the Vatican retable108. However, new technical analysis has confirmed that the Vatican altarpiece was always single-sided109. As we saw above, the fragments of Trecento frescoes discovered on the lower walls of the apse during the post-1997 conservation campaign suggest that the friars’ choir at Montefalco was in front of the high altar prior to the 1450s. It is unclear how long the Maestro di Fossa’s polyptych remained over the high altar at Montefalco, where its Passion scenes could in time have been complemented by Gozzoli’s Franciscan cycle. By the 1510s testamentary bequests indicate that a new tabula or ancona was at least envisaged for the high altar, but its format and iconography are unknown110.

29The evidence for single-sided high altarpieces becomes more plentiful as we move into the Renaissance period. In 1466-67 the Franciscans of Santi Simone e Giuda in Spoleto commissioned a high altarpiece from the local artist Jacopo Vincioli. According to the terms of the contract discovered by Silvestro Nessi and the surviving fragment in the Rocca Albornoziana this was a conventional single-sided pentaptych (fig. 40)111. Nothing is known of the church’s choir arrangement and the building’s interior has been altered beyond recognition112. Piermatteo d’Amelia’s single-sided high altarpiece for San Francesco, Terni of 1483-85 survives in excellent condition (fig. 41)113. The documentation for the Terni foundation is sparse but Piermatteo’s pala was never intended to be seen from the rear to judge from the bozze that are still visible on the back of the altarpiece114. Two Franciscan high altarpieces commissioned in 1506 from Pinturicchio for Sant’Andrea, Spello (fig. 42) and from Berto di Giovanni for San Francesco, Montone, seem to have been straightforwardly single-sided (Pinturicchio was also asked to incorporate a tabernacle into his pala, either in the predella or above it)115. As we have seen, Montone probably possessed a retro-choir by this date but the situation for the other churches is unclear116.

Photo: Roberto Sigismondi and Museo del Ducato di Spoleto

Photo: Roberto Sigismondi and Comune di Terni

30The accumulated evidence for high altarpieces in the Provincia Sancti Francisci points to a variety of practice from the beginning. The Maestro di San Francesco retable for San Francesco al Prato in Perugia establishes the double-sided tradition as early as 1272, reinforced at nearby Monteripido c. 1322. Meanwhile, the Maestro di Fossa’s single-sided altarpiece in the Vatican from San Francesco, Montefalco is securely dated by its donor inscription to 1336. The magnificent sequence of double-sided polyptychs at Perugia (1403), Città di Castello (c. 1390-1410) and Sansepolcro (1426-44) is balanced by single-sided polyptychs at Spoleto (1466-67) and Gualdo Tadino (1470). Similarly, Perugino’s double-sided Renaissance pala for Monteripido (1502) was matched by single-sided commissions for Piermatteo d’Amelia at Terni (1483-85), Berto di Giovanni at Montone and Pinturicchio at Spello (both 1506). Nor is there a tidy correlation between the format of altarpieces and choir configurations. The unpainted verso of Niccolò Alunno’s single-sided polyptych at Gualdo Tadino faced the church’s apsidal choir, although the altarpiece’s open arcade would have afforded the friars a glimpse of the high altar liturgy on the far side of the panel. In similar light, the friars at Sansepolcro asked Antonio d’Anghiari to leave the back of a double-sided polyptych unpainted in 1430 and the same option was entertained at Monteripido in 1452117. As Machtelt Brüggen Israëls has shown, unpainted panels were a frequent sight over Italian altars118. While the examples she cites were usually the result of incomplete projects or time-lapses between carpentary and artistic commissions, the Gualdo case suggests that an unpainted verso was an acceptable long-term solution. It is hard to believe, however, that Piermatteo d’Amelia’s study of a male nude on the back of the Terni high altarpiece would have been appropriate viewing for a Franciscan community; here at least the artwork itself provides evidence for a nave choir.

Umbrian retrochoirs beyond the Franciscan Order

31The foregoing consideration of retro-choirs and double-sided altarpieces suggests a more varied landscape of choir-altar configurations within the Provincia Sancti Francisci than hitherto thought, but there is yet another dimension that needs to be addressed for an adequate understanding of the retro-choir phenomenon in Umbria: the seepage of the tradition beyond the confines of the Franciscan Order to churches belonging to the secular and other branches of the regular clergy.

32The interior fabric of the Benedictine abbey of San Pietro, Perugia has long been associated with the city’s Franciscan churches thanks to Meo da Siena’s double-sided retable painted for its high altar in 1330 or 1333 (now in the Städel in Frankfurt)119. Following the exemplary reconstruction of San Pietro’s successive reconfigurations by Michael Gromotka, it is clear that the church possessed a retro-choir in the Trecento, which was probably related not only to Meo’s retable but also to the monumental polygonal apse added to the Romanesque basilica in the early fourteenth century120. The towering aspect of the new apse rivalled the Upper Church at Assisi on the opposite side of the Tiber valley. The medieval retro-choir was sacrificed after the Perugian abbey joined the Congregation of Santa Giustina in 1436, a network of Benedictine foundations across Italy that seems to have preferred nave choirs for its churches121. The locations of choir and altar in San Pietro were duly reversed, the former moved forward into the nave, the latter shifted back into the apse122. When the monks approached Perugino for a new high altarpiece in 1495, they commissioned a large single-sided pala suitable for a conventional nave choir arrangement123. This remained over the high altar until it was replaced by a Eucharistic tabernacle in 1567 and moved to a more elevated location on the rear wall of the apse124. The monastic choir was only transferred back into the apse in 1591, at which point the altar block was moved forwards to its present location125.

33Perugino’s other great high altarpiece project in Perugia for the Eremitani church of Sant’Agostino was, however, certainly double-sided126. The enormous wooden “tavola” for Perugino’s Augustinian altarpiece had been commissioned from Mattia di Tommaso da Reggio in 1495. The support was to incorporate a ciborium and be finely finished on both sides127. The friars’ contract with Perugino does not survive but according to the convent’s chronicle the artist agreed to paint the wooden support in 1502 for the fee of 500 gold ducats128. The renovation of the high altar also entailed the renewal of the conventual choir. In 1501 the Florentine architect and woodcarver Baccio d’Agnolo promised to submit two model stalls for the consideration of the Augustinian friars129. These evidently met with the community’s approval and the following year Baccio signed a contract for the entire choir for the grand sum of 1,110 florins130. The choir project, however, dragged on for decades and the Augustinians were still chasing Baccio for its completion in 1532131. The painting of the altarpiece itself took nearly as long: Perugino is thought to have finished the panels on the front between 1508 and 1512 but laboured on the back until his death in 1523132. In his 1568 edition, Vasari described Perugino’s “tavola grande isolata e con ricco ornamento intorno” painted both “nella parte dinanzi” and “di dietro, cioè dalla banda che risponde in coro133”. The altarpiece remained over the high altar until 1654 when it was removed and its two sides separated; in 1683 its principal elements were reset within stucco frames on the walls of the apse (fig. 43)134. What Perugino’s great pala had replaced remains unclear as we know very little about Sant’Agostino’s earlier arrangement and building history (fig. 44). The large polygonal apse that would prove capacious enough for Perugino’s high altar and Baccio’s retro-choir was under construction around the middle of the fifteenth century but formed part of a larger plan of expansion encompassing the crossing, transept and surrounding chapels that was begun much earlier in the fourteenth135. Further information is provided by the description of the church composed by Padre Giacomo Giappessi (d. 1720) in the early eighteenth century and drawn from the convent’s own archive, then extant but now almost entirely lost. Giappessi records that the concession of the cappella maggiore to Count Boncambio Boncambi in 1436 allowed the friars to construct a new tribuna behind the chapel and to install a choir there136. Describing the cappella maggiore as it then stood, the concession did not refer to a choir, from which we may tentatively deduce that the friars' stalls were not located behind the high altar before 1436137.

Reconstruction displayed in *Perugino: il divin pittore*, Galleria Nazionale dell’Umbria, 2004

34Evidence for other churches belonging to the Order of Augustinian Hermits in the region is mixed. James Banker has shown how in 1454 the Eremitani in Sansepolcro, having inherited an unpainted double-sided altarpiece from their Franciscan neighbours, asked Piero della Francesca to paint the front face only and not the back138. Ottaviano Nelli had painted a single-sided high altarpiece for the Augustinians at Pietralunga in 1403 and the same artist’s frescoes in the cappella maggiore of Sant’Agostino at Gubbio from c. 1420 leave little space for choir stalls below139. On the other hand, the Order’s church of Sant’Agostino in Montefalco possesses a generously sized polygonal apse with corbelled shafts that Alessandro Delpriori has suggested could have been designed for a retro-choir (fig. 45)140.

35A 1400 contract discovered by James Banker for the choir of the Servite church in Sansepolcro stipulated that the stalls were to be located post altare maius in the same manner as the recently completed choir in the local church of San Francesco. This modo et forma clause at Sansepolcro offers a particularly clear instance of liturgical arrangements favoured by one order influencing the choices made by another. The openness of the Sansepolcro Servites to a retro-choir arrangement may be relevant for the adventurous rotunda scheme planned for the Order’s church of Santissima Annunziata in Florence in the 1440s141. At Santissima Annunziata the liturgical arrangement of the rotunda was imposed by lay patrons despite the objections of the local Servite community, but the Sansepolcro evidence suggests that other churches belonging to the Order were also early-adopters of the retrochoir142.

36The situation for the region’s Dominican foundations is less clear. The Preaching Friars in Perugia commissioned a choir in 1476 for their new church of San Domenico; work on these stalls was still ongoing in 1498. They are now located in the church’s rectangular cappella maggiore but the Quattrocento documents do not specify their original location143. Spinello Aretino painted an impressive but single-sided altarpiece for San Domenico in Città di Castello around 1400 (fig. 46)144. The church has a rectangular apse (fig. 47) which now houses a set of choir stalls signed and dated OPUS MANNI DE FLORENTIA 1435 by the Florentine woodworker Manno di Benincasa, but it is unclear if these seats are in their original location. The Sienese artist Benedetto di Bindo painted a single-sided pentaptych for the high altar of Santi Domenico e Giacomo in Bevagna in 1416, while Gail Solberg has recently reconstructed another single-sided polyptych painted for the Dominican church in Gubbio by Taddeo di Bartolo in 1418145. In Spoleto, the frescoes discovered in the wake of the 1997 earthquake on the apsidal walls of San Domenico extend down to floor level and probably preclude the presence of choir stalls behind the high altar146. The Dominicans seem to have eschewed the regional tradition of polygonal apses with the exception of their church in Foligno, which had an expansive polygonal tribuna added to its single nave in the fifteenth century147.

Reconstruction by Silvia De Luca and Giorgio Guazzini

37This brief survey of the major religious orders can only scratch the surface of the enormous patrimony of ecclesiastical architecture in Umbria. For one thing, the abbey of San Pietro in Perugia – the only Cassinese house in the region – cannot be taken as normative for Umbria’s numerous Benedictine foundations148. Even within Perugia, the Silvestrine Benedictines of Santa Maria Nuova took a very different path from their Cassinese cousins at San Pietro, following the example of the local Franciscans instead by choosing a retro-choir and considering a double-sided high altarpiece. When commissioning a new choir in 1453, the Silvestrines specifically required that the 27 stalls should be located in the church’s polygonal apse, or in tribuna dicte ecclesie videlicet circum circha dictam tribunam149. The fine set of choir stalls signed by Paolino di Giovanni d’Ascoli and Giovanni da Montelparo in 1456 remains in the polygonal apse of Santa Maria Nuova today150. In 1465 the Silvestrines sought funding from the communal authorities to commission a high altarpiece, arguing that “there is no picture such as is requisite according to the importance of the said church151”. The panel support was evidently fashioned but not painted, and in May 1472 the Silvestrines sought further subsidies “since at the high altar […] there is a certain panel not painted, and since it has remained there for several years, and since by reason of want of opportunity […] and the poverty of the said convent, it could not be painted152”. In December that year they gave detailed instructions to the local painter Fiorenzo di Lorenzo for a double-sided high altarpiece153. The monks seem to have changed their minds when renewing the commission in 1487: they now stipulated that Fiorenzo should only paint one side of the panel, resulting in the single-sided polyptych which largely survives in the Galleria Nazionale dell’Umbria154. The reasons may have been more economic than liturgical, as the Silvestrines had pleaded poverty on more than one occasion when petitioning the Perugian authorities.

38An adequate treatment of the churches of the secular clergy lies far beyond the scope of this article but there is evidence of cathedrals in the region possessing retro-choirs before 1500. The 1349 contract for a new set of choir stalls for the cathedral of San Rufino in Assisi is especially valuable, not only for its early date but also for its specification that the new stalls were to be located in the church’s tribuna or main apse and that the seats were to imitate in size and workmanship those then in the Upper Church of San Francesco155. Niccolò Alunno’s 1483 high altarpiece for the cathedral at Nocera Umbra (fig. 48) reprised the open arcade that the artist had used for his 1470 high altarpiece for San Francesco, Gualdo Tadino, a feature – as we have seen – that may well have been intended to render the liturgy of the altar visible to an audience seated in a retro-choir156. The Perugian cathedral of San Lorenzo was rebuilt with a seven-segment apse in the fifteenth century and the choir stalls signed by Giuliano da Maiano and Domenico del Tasso in 1491 that are still in the apse seem to have been designed for that location157.

Conclusions

39This survey of Umbrian ecclesiastical fabric, already long, could easily be extended many times over. No mention has been made, for example, of collegiate, parish or nunnery churches. Even with regard to one relatively defined question – the location of choir stalls in relation to high altars within a single region from the thirteenth century to the sixteenth – the available material is vast and varied. The picture will doubtless continue to evolve as more evidence comes to light: we can, for example, expect further documentation pertaining to the installation or removal of choir stalls to emerge from Umbria’s still underresearched notarial archives. The foregoing review, however, does allow for some provisional conclusions.

40The first is that the early existence of retrochoirs in the region’s Franciscan churches can be verified with greater confidence than before through a combination of archival sources and surviving fabric. The documented case of Sansepolcro pushes the phenomenon back to the second half of the Trecento while the surviving stalls at Gualdo Tadino are likely to have been carved at the start of the century. If the high altar arrangement in San Fortunato, Todi dismantled in 1580 did indeed date back to 1301, then that church was probably built to house a retro-choir in the 1290s. Even earlier, combined factors suggest that San Francesco al Prato in Perugia was designed to accommodate a retro-choir in the middle of the thirteenth century. Alongside this “proof” of the retrochoir hypothesis, however, comes the realization that other foundations within the Provincia Sancti Francisci adopted different configurations. Evidence for this comes in the form of Trecento frescoes that may preclude the presence of apsidal choir stalls at Montefalco, Montone, Trevi and Vallo di Nera. The situation is complicated by the likelihood that the churches at Montefalco and Montone later seem to have been retrofitted with retro-choirs in the mid fifteenth and early sixteenth centuries respectively.

41The picture is muddied further by the clarification that features sometimes associated with retro-choirs are not necessarily reliable markers. Corbelled vaults facilitated the installation of stalls in the apses at Gualdo Tadino and Cascia but similarly truncated shafts at Montefalco and Montone were not associated – at least initially – with retro-choirs. These variations invite us to reconsider the consistency and degree of coordination between architectural design and liturgical configuration. The situation with high altarpieces is likewise less uniform than hitherto thought. For now, all known instances of Franciscan double-sided altarpieces in the region – at Sansepolcro, Città di Castello and the two Perugian foundations of San Francesco al Prato and Monteripido – can be linked to churches for which there is documentary evidence for a retro-choir. But double-sided altarpieces are outnumbered by the single-sided images commissioned for Franciscan high altars across Umbria. Some of these seem to have shielded retro-choirs (as at Gualdo Tadino) while others must have faced conventional nave choir arrangements.

42If the adoption of retro-choirs amongst the Umbrian Franciscans was not consistent, nor was it exclusive to them. In the region’s principal urban centre, Perugia, the retro-choir was arguably a civic phenomenon rather than a specifically Franciscan one158. The city’s major Benedictine foundation, San Pietro, was equipped with a retro-choir in the early fourteenth century, although this arrangement was later reversed in the fifteenth. Retro-choirs can be identified in several more of Perugia’s larger churches by the end of the Quattrocento: the mendicant house of Sant’Agostino, the Benedictine church of Santa Maria Nuova and the city’s cathedral of San Lorenzo. Elsewhere, the Upper Church at Assisi inspired a Trecento retro-choir in the local cathedral, San Rufino, while the arrangement of San Francesco in Sansepolcro informed the commission of an apsidal choir by the town’s Servite community in 1400. At Nocera Umbra, Niccolò Alunno’s 1483 high altarpiece for the cathedral reprised the perforated polyptych design that the artist had earlier devised for the retro-choir arrangement in San Francesco, Gualdo Tadino in 1470.

43Nor was this process of seepage confined to Umbria’s borders. The Franciscan retro-choir in Sansepolcro was cited as a modo et forma model in 1400 not only by the local Servites but also their confrères from Sant’Angelo in Vado over the Apennines in the Marche159. Meo da Siena’s dossal for San Pietro, Perugia soon inspired a similarly double-sided altarpiece for the ancient Benedictine abbey at Subiaco in Lazio160. The later history of San Pietro also shows how connections with other parts of Italy could work in reverse, with the Trecento retro-choir being removed to make the Perugian abbey conform to Cassinese practice pioneered in North-Eastern Italy (see Gromotka’s essay in this book). While the popularity of retro-choirs amongst the Umbrian Franciscans probably reflected the administrative structure of the Order’s Provincia Sancti Francisci, such regional limitations did not apply to other branches of the secular or regular clergy161.

44Major questions remain regarding the liturgical and ritual practicalities of retro-choir arrangements. How did religious communities in apsidal choirs participate in the liturgy of the high altar, from which they may have been screened by retables and polyptychs? Such evidence as there is indicates that Mass in such cases would have been celebrated on the “lay” side of the altar mensa: i.e. while altarpieces could be double-sided, altar tables were not. Apertures, arcades and tabernacles of various kinds may have facilitated visual access from behind: the detailed instructions given by the Perugian Silvestrines to Fiorenzo di Lorenzo in 1472 for a double-sided altarpiece foresaw a “finestre che se vede el corpo de Cristo162”. Circumstantial evidence suggests that similar “windows” or openings may have been set within the predellas of other double-sided high altarpieces163. Later Cinquecento arrangements may provide clues for earlier practice. Birgit Laschke has drawn attention to the well-preserved and well-documented double-sided high altarpiece executed by the Servite sculptor Fra Giovanni Angelo da Montorsoli for his order’s church in Bologna between 1558 and 1562164. Montorsoli’s commission, like Giorgio Vasari’s slightly later high altarpiece for the Pieve in Arezzo (1560-64), stands on the cusp of liturgical change, responding to the Tridentine emphasis on Eucharistic display but still subordinating it to a larger, free-standing altarpiece structure165. Montorsoli’s design incorporated a Eucharistic tabernacle within the thickness of the predella which was covered with a relief-image of a chalice adored by angels on the front and closed with an open grate at the back to facilitate veneration from the conventual choir166.

45The degree of lay access to retro-choirs also remains unclear. Both Machtelt Brüggen Israëls and I have argued from notarial sources that the laity could gain periodic access to choir precincts and other conventual areas, but nor should we underestimate the need for seclusion on the part of mendicant and monastic communities: this figures as a key concern in Giovanni Aldobrandini’s commentary on the rotunda retro-choir envisaged at Santissima Annunziata in Florence167. In Umbria the earliest double-sided retables would hardly have obscured apsidal choirs by themselves, but the desire to do so may have encouraged the development of larger polyptychs and pala altarpieces that would have served as visual barriers. Moreover, it appears that double-sided altarpieces did not render tramezzo screens redundant. Evidence for nave screens – although nowhere definitive – can be accumulated from Città di Castello, Gubbio, Montefalco, Sansepolcro, and Trevi; enough to suggest that tramezzi were common features in the region’s Franciscan churches regardless of whether choir stalls were arranged before or behind high altars. This in itself marks a divergence from the model of the Upper Church, which seems to have been conspicuously open-plan at least from the 1290s onwards168.

46What purpose would the division of the nave have served if the choir was in the apse? We know that in Città di Castello in 1492 the lower nave was reserved for the female laity. Bram Kempers has recently argued that the documentation for Sansepolcro records not only a retro-choir but also implies a second choir precinct in the upper nave169. There are later analogies for such double choir arrangements in other regions: returning to the case of Santa Maria dei Servi in Bologna, this church was equipped with a retro-choir for its ordained friars in 1450 while the lay brothers continued to use the nave choir, a situation that persisted until the removal of the latter in 1611, long after the nave tramezzo was demolished in 1563170. References cited above for the Franciscan churches in Montefalco and Bevagna to a corus magnus may imply the contemporaneous existence of a spatially separate corus minor, or may simply refer to the upper of two banks of choir stalls in a single precinct171. Future documentary finds will hopefully clarify the question, but again we are reminded of Sible de Blaauw’s observation that the early history of the retro-choir was not tied to a desire to clear the nave for the laity.

47While the distinctive nature of early Franciscan sacred interiors in Umbria can now be verified with greater confidence for a core group of buildings, the overall situation across the region is clearly more complex and varied. Inconsistent Franciscan practice and the ease with which retro-choir arrangements were adopted by other ecclesiastical patrons suggest considerable latitude in the spatial configuration and physical provisioning of the Latin rite, with ramifications for our tendency to characterize the liturgy as a functional framework for images and artefacts. In a richly textured landscape of intersecting patterns different choices were evidently available but we still struggle to discern the connotations or traditions that informed them. Were retro-choirs considered an ancient or modern feature? Did they carry papal echoes mediated by the regional model of the Upper Church, even if the retro-choir in San Francesco al Prato proved ill-equipped to host the papal liturgy when the need arose in 1506? The sources offer only fleeting hints and any analysis is hindered by the very vague understanding we have of Umbrian traditions of liturgical layout prior to the thirteenth century172. For the moment, the most plausible explanation for the Umbrian retro-choir is the authority – aesthetic as much as liturgical – of the Upper Church. But much work remains to be done – for Umbria as elsewhere – in what may now be considered a distinct field of enquiry evolving its own methodology: the study of the church interior, its physical fabric, performative qualities and visual affordances.

Notes de bas de page

1 Cooper 2001.

2 The key contribution here being Os 1974. See esp. p. 121-122 for the author’s analysis of the layout of San Francesco al Prato, Perugia and Taddeo di Bartolo’s double-sided polyptych painted for that church in 1403: “One of the most characteristic architectonic features of the original church was the pronounced apse. This is where, in Franciscan churches, the monks’ choir was located, and it is this part of the church which we must link with Taddeo di Bartolo’s […] high altarpiece, which must have divided the choir from the main body of the church”. Van Os did note that many churches, including Franciscan buildings, separated laity from the clergy “in a completely different way […] by means of a rood screen in the nave”, but concluded that “there is no doubt, however, that an altarpiece such as Taddeo di Bartolo’s shows that the laity had become directly involved in the liturgical center and, secondly, that the monks practiced their devotions behind that center”. This assertion was qualified by “the fact that very little has been published on the original liturgical use of Franciscan churches and its influence on their architectural form”.

3 Writing several years after Van Os’s article but without reference to it, James Ackerman judged the transposition of altar and choir “a revolution in monastic church planning and liturgical practice as radical as any in the history of monasticism” but identified the earliest example of such a reconfiguration to be Michelozzo’s rebuilding of Santissima Annunziata in Florence initiated in 1444 (Ackerman 1980, p. 294-295). For other surveys of early retrochoir arrangements (again with no mention of Umbrian material), see Isermeyer 1950, p. 140-141, note 15 (identifing Cosimo de’ Medici’s reconfiguration of Brunelleschi’s plan for San Lorenzo after 1442 as the oldest known example, inspired – according to Isermeyer – by exchanges with Greek prelates at the Council of Ferrara-Florence); Clearfield 1981, p. 25-27 (again emphasizing San Lorenzo, speculating that the trend was a revival of Early Christian practice); Struchholz 1995, p. 103-110 (with a similar range of Renaissance examples, but noting the presence of choir stalls behind as well as in front of the high altar of Siena cathedral at the end of the Trecento). Brown 1996, p. 175, argued that Michelozzo’s Santissima Annunziata scheme “spearheaded a mini-revolution in the reorganization of Florentine church interiors”. For initial comments on the distinct but related tradition of the so-called “Mausoleum Choir”, where tombs of eminent lay and ecclesiastical patrons were placed in axial proximity to high altars with surrounding choir stalls, see Frommel 1977, p. 46-51.

4 Cooper 2001, p. 54.

5 See the conference report by Romagnoli 2003, p. 307-308. In the published proceedings, see the introduction by Jörg Stabenow (Stabenow 2006, p. 11), and the extensive discussion of my 2001 article in the chapter by Sible de Blaauw (Blaauw 2006, p. 39-43).

6 Bacci 2005, p. 117; Hall 2006, p. 218.

7 Here the publications of Joanne Allen are particularly significant (see Allen 2013b and Allen 2013c). Haude Morvan has shed much new light on the adoption of retrochoirs by the Dominicans (see Morvan 2013 and 2021). For early retrochoirs in Rome, see Blaauw 2006, p. 43-49. The installation of a retrochoir in San Lorenzo – often advanced as the first Florentine example – has now been dated later to c. 1480 by Riccardo Pacciani (Pacciani 2017).

8 Perhaps for this reason, Sible de Blaauw (Blaauw 2006, p. 39, note 31) doubted my argument that retrochoirs could be identified in Umbria as early as the thirteenth century (“Credo tuttavia audace la conclusione secondo la quale il fenomeno si sarebbe stabilizzato dal 1250 in poi”).

9 For a survey of the field, see Cooper 2017.

10 On the Arezzo plan, see Borgherini 2001, p. 164-183; Cooper 2001, p. 6-7, and more recently Cooper 2011, p. 93-96.

11 Although, as Sible de Blaauw observes (Blaauw 2006, p. 34), the church’s apse points to the south-west, not eastwards. The step drawn in front of the high altar in the Arezzo plan confirms that the priest would have celebrated mass with his back turned to the congregation.

12 On the administrative structure of the province in the early fourteenth century, see Pellegrini 1984, p. 300-301. The total number of 74 foundations, organized into nine custodies, includes remote mountain hermitages as well as groups of friars attached to major Clarissan nunneries like Monteluce in Perugia. Pellegrini’s list is based on the Provinciale Ordinis Fratrum Minorum compiled by Fra Paolino da Venezia, probably around 1330. The division of the province into nine custodies (secundum novenarium numerum) is already recorded in its chapter acts of 1316, see Cenci 1963. The Provincia Sancti Francisci broadly corresponds to the modern political region of Umbria, but two Umbrian centres along its western border, Città della Pieve and Orvieto, fell into the Order’s Provincia Tuscie and Provincia Romana respectively. Conversely the Provincia Sancti Francisci spilled out of Umbria to include Sansepolcro in Tuscany and six convents in the mountain valleys to the south and east of Norcia that now lie in the Marche or Lazio. The most valuable catalogue of mendicant architecture in the region is Curuni 1982, with earlier surveys by Biebrach 1908 and Krönig 1938 comparing mendicant buildings in Tuscany and Umbria. The architectural scholarship is generally taxonomic, tracing typologies in groundplans, vaulting systems, and the degree to which Umbrian churches reprise formal aspects of the Upper Church at Assisi, see Schenkluhn 2000, p. 56-63. Surveys rarely consider liturgical furniture as an integral part of a building’s design: see the important historiographical observations by Bruzelius 2014, p. 1-17.

13 Cooper 2001, p. 32-39, 52, posited the Upper Church as a model not only for the apsidal architecture of its Umbrian daughter houses, but also for their retro-choir arrangements. The vaults of the Upper Church were certainly standing by 1253, when they are mentioned in an account of the canonization of Saint Stanislaus staged in the space on 17 September that year. The early presence of choir stalls in the apse is indicated by the curtailed lower edges of Cimabue’s frescoes, probably painted by c. 1280. For further discussion, see Cooper – Robson 2013, p. 62-63. However, Blaauw 2006, p. 39-41, has noted that apsidal seating in papal arrangements in Rome served very different liturgical purposes to conventual choir stalls.

14 Israëls 2009.

15 Cooper – Banker 2009, see esp. p. 69-76, with documentary appendix (Israëls 2009, II, p. 585-589).

16 Israëls 2009, II, p. 586, doc. 4. This contract, dated 14 May 1378, also specified that the choir had been in the same location before the earthquake: qui ibi esse solebat ante terremotos quorum vi devastatus fuit. For the initial transfer of 100 florins from the fraternity of San Bartolomeo on 21 October 1376 “convertendose in lo refare et reconciare del coro d’essa ecclesia”, see Israëls 2009, II, p. 585, doc. 2.

17 See Israëls 2009, II, p. 587, doc. 6, for the agreement of 24 June 1392 between the Operai of the church and the woodcarver brothers Giovanni and Angelo Schiatti, who agreed to make “nel decto cuoro trenta sedie, cioè XV per parte, remanendo la sedia grande nel mezzo del decto cuoro et debbono fare el lavorio de le decte sedie in quella forma et a quella similitudine de le sedie del cuoro menore” (the reference to the lesser choir indicates that this agreement concerned the upper tier of stalls). A set of surviving choir stalls in the Franciscan church in Zadar (Croatia) are documented as the work of one Giovanni da Sansepolcro via a payment of 1394. However, documents discovered in Venice clarify that this could not have been the same Giovanni that worked on the choir in San Francesco, Sansepolcro, see Schulz 2011, p. 117; Cooper 2014, p. 61, note 41.

18 For the likely existence of a screen prior to the Apostolic Visitation conducted by Angelo Peruzzi in July 1583, see Cooper – Banker 2009, p. 98-103 (and for a groundplan, see fig. 20 on p. 57).

19 Blaauw 2006, p. 42: “D’altronde, non pare che sia stato il desiderio di offrire una maggiore visibilità dalla navata verso il santuario a costituire un argomento consistente per lo spostamento del coro oltre l’altare, se non dal Quattrocento avanzato. Questo è testimoniato proprio dai casi in cui il coro fu trasferito dietro l’altare senza che il tramezzo venisse smantellato”. Bram Kempers has read the Sansepolcro documents differently to mean that “probably two choirs co-existed on both sides of the high altar, one in front of the high altar and the other at its rear side. […] During Mass the friars had the first choir at their disposal, viewing the front side of Sassetta’s polyptych; at other occasions, during Divine Office, they viewed the back side, using the retrochoir” (Kempers 2016, esp. p. 276-277 and note 11). See also note 169 below.

20 In 1400 the completed stalls in San Francesco, described as novi facti, were cited as a modo et forma model in two contracts for choir precincts in the Servite churches in Sansepolcro and Sant’Angelo in Vado, see Cooper – Banker 2009, p. 72, and docs 8-10 (Israëls 2009, II, p. 587-588).

21 The detailed documentation for the commission was discovered by James Banker and published in his seminal article: Banker 1991; Banker re-edited the sources as Appendix of Documents Relating to the High Altarpiece, the High Altar, and the Tomb of the Blessed Ranieri in the Church of San Francesco in Borgo San Sepolcro, in Israëls 2009, II, p. 566-583.

22 At over 10 metres wide, the cappella maggiore at Sansepolcro is proportionately larger in relation to the church than the rectangular high altar chapels of many Franciscan churches in Tuscany. In addition to the main window on the eastern wall, the chapel was lit by single lancet windows on the side walls, a feature that is peculiar to Sansepolcro and would have introduced much more daylight into the space, see Cooper – Banker 2009, p. 62-63. Assuming that the post-1352 rebuilding work on the vaults did not significantly alter the building’s groundplan, these features indicate that the church was designed from the start to accommodate a retro-choir. The site had been granted to the Franciscans as early as 1258 (Israëls 2009, II, p. 585, doc. 1) and the high altar mensa bears the date 1304.

23 Although the vault was demolished to make way for a new apse in the eighteenth century, its simple faceted corbels and the lower portions of its shafts are preserved in the intermediate space between the old and new apses, see Cooper – Banker 2009, p. 62. The arrangement of the cycle of scenes from the life of Saint Catherine of Alexandria frescoed in the cappella maggiore in 1403 also allows space for the choir stalls below, see Cooper – Banker 2009, p. 76-84; our reconstruction was subsequently refined and revised by Squillantini – Guazzini 2012.

24 Schenkluhn 2000, p. 56-59.

25 For the stalls at Cascia and Gualdo Tadino, see Cooper 2001, p. 27-32. On the architecture and history of the Gualdo church, consecrated in 1315, see now Lunghi 2019 (I am grateful to the author for the opportunity to read his essay prior to publication).

26 On 15 April 1449, Guido di Pierpaolo Petrucci left 200 lire for duabus fenestris de vitro fiendis et construendis ponendis videlicet in fenestris existentis in capella cori et in maior capella dicte ecclesie [San Francesco] (ASCCC, Archivio Notarile, 18/7, fol. 155v). Guido’s bequest was honoured by his daughter Bianca on 7 January 1454, who paid the promised sum pro fabrica sive aconcimine duarum fenestrarum de vetro in ecclesia predicta supra corum dicte ecclesie Sancti Francisci a parte posteriori altaris maioris dicte ecclesie (ASCCC, Archivio Notarile, 26/3, fol. 118v). These documents were first published in Cooper 2016 at p. 110, note 33. The interior of the medieval apse is now buried within the thickness of the smaller apse inserted in the eighteenth century and it is impossible to tell whether its vault was corbelled in a similar fashion to Cascia and Gualdo Tadino, see Curuni 1982, p. 95-98 (cat. 5.5).

27 On 1 April 1486 the friars at Città di Castello sold property pro fabrica cori fiendi in dicta ecclesia et aliis oportunitatibus dicti conventus (ASCCC, Archivio Notarile, 29/8, fol. 245v). The new choir had been envisaged for some time; on 26 June 1475, Ser Luca Albizzini had left 200 gold florins pro fabrica unius cori fiendi in dicta ecclesia Sancti Francisci should his children die without legitimate heirs (ASCCC, Archivio Notarile, 29/6, fol. 231v).

28 There are cases where pulpitum and pergulum unequivocally served as synonyms for screen structures, see Cooper 2001, p. 47, for the clear-cut instances of the Franciscan churches in Perugia and Bologna. However, see De Marchi 2008, esp. p. 389-391, for the etymology and evolution of the term pergulum between the fourteenth and fifteenth centuries. De Marchi argues persuasively that while the word usually referred to a screen in the fourteenth century, by the Quattrocento it could also refer to the separate category of isolated preaching pulpit located in the lower nave.

29 In his testament of 31 July 1413 Giovanni Salvini from Città di Castello asked quod de bonis suis fiat una capella in muro eglesie Sancti Francisci supra capellam filiorum Nicolai Nofrii versus pergulum predicatoris (ASCCC, Archivio Notarile, 11/11, fol. 19r). A notarial act of 3 September 1458 was rogated in ecclesia Sancti Francisci infra pulpitum et portam magnam dicte ecclesie iusta plateolam (ASCCC, Archivio Notarile, 26/7, fol. 42v). On 18 April 1492, Giovanni Pietro Buffalini was granted space in San Francesco for unum altare cum capella in pariete eclesie dicte domus dividente eclesiam ipsam a claustro eiusdem domus in parte inferiori dicte eclesie deputata pro residentia mulierum inter penultimam et ultimam ex tribus capellis sitis in dicta eclesia ex parte inferiori iuxta dictum parietem (ASCCC, Archivio Notarile, 32/1, fol. 88v-89r, cited with abbreviated transcription in Cooper 2001, p. 45).

30 Cooper 2001, p. 32.

31 See Nessi 2002, p. 26-27, for three separate testamentary bequests of 10 florins pro coro magno fiendo in dicta ecclesia (5 September 1455), 9 florins pro coro fiendo in dicta ecclesia (21 January 1451), and 6 florins pro coro fiendo in dicta ecclesia (12 October 1457).

32 In 1336 the friars had vacated their earlier convent of Santi Filippo e Giacomo outside the walls of Montefalco in order to move to a more secure location inside the town. The new church of San Francesco was receiving legacies for Masses by 1342 and was being used for court hearings in 1344. The earliest evidence for wall paintings comes via a testamentary legacy of 1363. For these excerpts from local notarial acts, see Nessi 2002, p. 14-16.

33 See the catalogue entry by Francesca Cristoferi in Toscano – Capitelli 2002, p. 297-300, fig. 12-13. The fragments beneath Gozzoli’s scene of the Homage of a Simple Man indicate an image of a Virgin and Child whose lower portions are likely hidden behind the present stalls (“di dimensioni tali da occupare tutto il primo lato del poligono absidale e di cui si potrebbero trovare altre tracce dietro i postergali del coro ligneo”).

34 See Nessi 2002, p. 60, for a testamentary bequest on 10 January 1515 pro lampada Crucifixi existentis in medio dicte ecclesie. Delpriori 2015, p. 251-252, has convincingly associated the elevated niche midway along the left nave wall with the access staircase for a pulpit platform on a tramezzo screen (see fig. VIII.46 for the reconstruction of the interior with the screen reinstated).

35 Gillerman 1987, p. 62, had already linked the corbelled vault shafts with the presence of an apsidal choir: “At S. Fortunato, however, the major colonnette rests on a corbel which projects from the wall ca. 4.50 m. above the present level of the floor. […] Presumably, the motivation for the new treatment of the apse colonnettes at S. Fortunato was the desire to allow projected choir stalls to be pushed up flush against the choir wall. Hence, Matteo da Gubbio’s late sixteenth-century choir stalls were probably commissioned to replace medieval originals”. For Maffei’s commission, see Gambella 2008, p. 117-120.

36 See De Angelis D’Ossat 1982, p. 168 and 169 for payments “per acconciatura de la finestra del vetro che sta sopre el coro” (28 November, 1432) and “ad rasattare et risarcire la vitriata sopra el coro” (31 December, 1556).

37 For general considerations about this kind of source, see Morvan’s chapter in this volume.

38 According to Giambattista Possevino (Possevino 1597, p. 120-121), the bodies of the three saints, together with the virgin martyrs Romana and Degna, were rediscovered below the high altar on 13 December 1580, “guastandosi detto altare, che occupava gran parte della detta chiesa […] con intentione di rifarlo in luogo piú commodo”.

39 Giovan Battista Guazzoroni, Collectanea historiarum Tudertinarum, Archivio storico comunale di Todi, Arm. VI, cas. I, no. 8, IV, fol. 230r: “Scipione Benedettonio e Giovanni Agostino degl’Atti rettori di Santo Fortunato di Todi fecero scarcare l’altare di detta chiesa per condurlo più vicino al choro, e trovarono sotto al detto altare una gran cassa di trevertino”; cited without the manuscript source by Grondona 1991, p. 144, and Cooper 2001, p. 27. Possevino 1597, p. 122, similarly noted the subsequent construction of “una Cappella sotto l’Altar maggiore, gia rifatto più verso il coro.”

40 Possevino 1597, p. 120-121: “I corpi delli santi […] si trova, che dell’anno 1301 alli XIX d’Agosto al tempo di Monsignor Nicolò Armato Vescovo di Todi, furono collocati sotto l’Altar maggiore della Chiesa di San Fortunato, dove sono stati sempre con molta divotione riveriti, fin’all’anno del 1580.”

41 Haude Morvan has identified a similar case from 1547 of a high altar being moved backwards not forwards in the Dominican church of Santa Maria in Gradi at Viterbo, although in this case the explanation is likely to have been the presence of the papal tomb-shrine of Clement IV in the medieval apse rather than an early retro-choir, see Morvan 2013, p. 101-02, fig. 2. Another instance is Santa Maria del Popolo in Rome, where the high altar was moved back towards the apse in 1624-27. As in the Franciscan churches in Todi and Perugia, this was a reconfiguration of a pre-existing retro-choir arrangement, dating at least from Bramante’s extension of the main apse of the Roman church in 1505-9 and perhaps already foreseen in the Sistine rebuilding of Santa Maria del Popolo in the 1470s, see Blaauw 1996 and Pöpper 2000, p. 258-60, fig. 3. I am grateful to Sible de Blaauw for bringing this parallel to my attention.

42 Cooper 2001, p. 9-17; Cooper – Sartore 2020.

43 Crispolti 1648, I, p. 139: “vedesi sotto la tribuna l’altare maggiore isolato, e fatto di marmi rossi, il quale è sacrato, come anche la chiesa, fin dall’anno 1536”; Cooper 2001, p. 14.

44 ASPg, Corporazioni religiose soppresse, San Francesco al Prato, Serie Miscellanea, 14, fol. 33r: “Sepolture novamente fatte dove prima era lo altare grande che sono in tutto 25.”

45 Cooper 2001, p. 13-14, fig. 8-9; for a slightly revised reconstruction, placing the pre-1536 high altar closer to the centre of the crossing, see Cooper – Sartore 2020, p. 78-80, fig. 18-19.

46 On 15 December 1530 the procurators and friars of San Francesco al Prato contracted Magister Iohannes Battista Francisci aliter Bastone to construct a choir within three years, calculated from January 1531. The contract specifies various formal and structural features of the stalls with reference to others in the local churches of Sant’Agostino and San Lorenzo, but does not provide any indication of the choir’s location within San Francesco al Prato. Giambattista Bastone’s sons Girolomo and Eusebio finally settled their father’s accounts with the friars for the choir on 5 April 1560, see Rossi 1873, p. 61-62, doc. 49. Rossi noted (p. 62) that “Lo stromento col quale i frati allogarono al Bastone l’opera del coro era per mano di Mariotto Calcina; ma io l’ho cercato invano”; this is the 1530 contract that has now been found by Alberto Sartore, to whom I am very grateful for sharing his discovery. For a preliminary consideration, see Cooper – Sartore 2020, p. 83.

47 See ASPg, Riformanze, 104, fol. 63r-v, for the payment by the Commune for reparatione […] fundamentis suptus corum et tribunam dicte ecclesie in July 1468. Payments in 1476 pro reficienda et reactanda fenestra vitrea chori ipsius ecclesie which had been tempestatem ventorum […] confracta may refer to windows of the main apse; see Riformanze, 112, fol. 113v, 117r. Cited by Cooper 2001, p. 16, and without transcriptions by Ricci 1927, p. 34.

48 For Innocent’s bull Petitio vestra, given at the Lateran in Rome on 19 October 1253, the pope recalled how in eodem loco, in quo Ecclesiam construere intenditis, primarium posuimus lapidem, see Bullarium franciscanum, I, p. 685-686. The ceremony could have taken place at any point during the pope’s Perugian sojourn from November 1251 until April 1253. Writing from Lyons on 14 June 1248, Innocent had already conceded 40 days of indulgence for those aiding the construction of a new church and convent, indicating that work on San Francesco al Prato was already underway or at least envisaged by that date (see Bullarium Franciscanum, I, p. 520). For San Francesco al Prato, see Borgnini 2011, esp. p. 13-17 for its early building history and p. 76-79 for the arrangement of the high altar and choir, and now Cooper – Sartore 2020.

49 The comparative measurements are taken from Gillerman 1987, p. 59-60, who identified Todi’s dependence on Perugia: “In this respect, the S. Fortunato apse depends on the apse of the church of S. Francesco al Prato at Perugia […] the first in Umbria to have assumed a seven-sided plan.”

50 Curuni 1982, p. 102-103 (cat. 5.7): the church received indulgences in 1256 but was still incomplete in 1291 to judge from a bull of that year conceding indulgences for those giving alms ad perficiendam ecclesiam Sancti Francisci Eugubii […] iam dudum incoeptam quam modo consummare incoeperunt opere plurimum sumptuoso.

51 For a description of these foundations (which were subsequently covered with the relaying of the floor), see Rosati 1983a, p. 7: “Lungo la navata maggiore sono poi emersi due importanti reperti. Il primo, al centro della zona della quarta campata, partendo dall’abside centrale: è un muretto in pietra, dello spessore di 45 cm. poggiante direttamente sul pietrisco originale. Il muretto partendo dalla terza colonna a sinistra, si distende verso la quarta per quasi due metri, e fatto un angolo retto, attraversa tutta la navata (centrale), quindi di simmetricamente si ricongiunge alla terza colonna di destra. Era il muretto che serviva da delimitazione e da basamento di sostegno del coro medioevale dei religiosi e si estendeva nella zona della seconda e terza campata, davanti all’altare maggiore che in quel tempo era collocato sotto le vele dell’abside centrale, come in altre chiese francescane (tipica attualmente quella di S. Maria Gloriosa dei Frari di Venezia)”. The plan reproduced here is taken from Rosati 1983b, p. 163 (plate 96).

52 Ser Balduccio Spadalonga’s testament of 27 January 1420 left 15 gold florins fratribus, capitulo et conventui loci Sancti Francisci de Eugubio pro duabus fenestris fiendis de vitrie in tribuna seu capella magna dicte ecclesie (Archivio di Stato di Gubbio, Fondo Armanni, I.G.22; cited by Rosati 1989, p. 67).

53 Lilli 2000, p. 138-139.

54 On the architecture of the Trevi church, see Curuni 1982, p. 136-138 (cat. 5.20), and Schenkluhn 2014, p. 67-88 (where the term “coro” is rather confusingly applied to the architectural superstructure of the building). Documents from the 1350s indicate significant building work, but Bruno Toscano’s recent catalogue entry on the apsidal frescoes argues that the apsidal end of the church was already standing by the 1340s, a chronology supported by the style of the frescoes and the casting of a bell for the campanile in 1341, see Toscano 2014, p. 144-152 (cat. 23).

55 See the entry by Delpriori 2016, p. 190, who notes that below the fresco cycle of Marian scenes in the main apse at Trevi “è una decorazione a finto marmo che doveva essere per forza visibile e che quindi, necessariamente, pone il coro in un’altra posizione, immaginiamo più avanti sulla navata”. However, we cannot discount the possibility that such decorative dado registers may have been disposable elements, especially in cases where choir stalls were foreseen but not yet ready (recalling the extended schedule documented for the choir commission at Sansepolcro). The inclusion of fictive marble panels or textiles below a narrative cycle would have been a swift and relatively inexpensive addition that would have avoided an untidy break along the lower edge of a fresco cycle of the kind that is still visible below Cimabue’s apsidal scenes in the Upper Church. I am grateful to Gail Solberg for her observations on this question.

56 For this bequest from the testament of Sensina daughter of Luca di Nerio of Foligno and wife of Luca di Lucarello of Trevi, dated 22 February 1433, see Marcelli – Felicetti 1999, p. 566-567 (doc. 10).

57 According to the contract agreed on 6 November 1480 the masons Baldassare di Giorgio and Antonio di Pietro, both from Como, would undertake extensive work on the walls and vaults of the sacristy, refectory and dormitory of the convent, including: Item quod in capite dicti andavini [a passageway that would run above the vault of the sacristy] teneatur facere hostium de lateribus et de inde scalam cum parapecto per quam nunc itur de ecclesia iuxta corum in inclaustrum. For a full transcription, Felicetti 2014, p. 291-292.

58 The convent was “ricostruito dalle fondamenta” in 1640-48, see Petrini 2014, p. 94.

59 Pirri 1921, p. 191: “1502 et addi XXII di febraio, la vigilia di santo Mactia, lu Reverendissimo Generale dell’ordine de santo Francisco de frati minori, nominato magistro Egidio da Melia homo de grande santità, intrò in Trevi et in santo Francisco con tanta devotione che me parse miravigliosa cosa. Andò all’altare della ecclesia de sancto Francisco et li disse sue oratione. Poi fo posto ad sedere in una sedia in dicta ecclesia, lato la intrata ad quella rata all’altare grande, et posto ad sedere, lu ministro della provincia se ingenochiò et basiò in viso ad quillo, et cusi sequitarono più frati […]. Stava in quella sedia tanto humano, gratioso, el viso suo pariva resplendesse de omne santità, con tanta gravità et suavità, me pariva fusse uno altro sancto Francisco.”

60 For San Francesco, San Gemini, a monument for which there are few documents but which is generally dated to the fourteenth century, see Curuni 1982, p. 126-127 (cat. 5.16).

61 For San Francesco, Narni, see Curuni 1982, p. 109-111 (cat. 5.10), who dates the five segment apsidal vault with corbelled shafts to 1324-37, when the local bishop was the Franciscan Amanzio. The apse is approximately 10 metres wide (equalling the width of the central nave) and 10 metres deep, and can be compared to the tribune of the nearby cathedral of San Giovenale, under construction from 1322. Vignoli 2000, concurs with Curini’s building history and dates (p. 76-77) some of the surviving frescoes in the nave and side chapels to the first half of the fourteenth century. The awkward manner in which the apse is joined to the earlier nave suggests that a cappella maggiore on this scale was not initially envisaged.

62 For the church, see Braconi 2013, who notes (p. 69) that in 1348 the building already required repairs and (p. 70-71) that the carving of the corbels in the apse recall those in the Palazzo dei Consoli in Gubbio, executed around 1322. For the Renaissance stalls, which seem to have been constructed for their present location, see Carla Mancini’s catalogue entry in Sapori 1997, p. 137, cat. 80.

63 For Antonio Alberti’s cycle, see Santi 1967, and the catalogue entry by Maria Rita Silvestrelli in Sapori 1997, p. 68-81, cat. 27.

64 For the consecration cross see Silvestrelli’s separate entry in Sapori 1997, p. 82, cat. 28.

65 See Felicetti 2001, p. 254, for testamentary bequests for the Franciscan church in Amelia pro picturis… [trebu]na (doc. 2: 26 April 1324) and pro picturis fiendis in trebuna (doc. 4: 27 March 1371). The cappella maggiore at Amelia was later frescoed by two painters from Narni, Pietro di Angelo and Cristoforo di Matteo between 1431 and 1433. On 5 September 1431, they agreed with the syndic of San Francesco to paint a chapel que cappella est cappella principalis dicte ecclesie Sancti Francisci, cum ymaginibus et signis secundum quod fuerit consignatum per fratres dicti loci, cum tribuna et parietibus ipsius capelle ab extra usque ad tectum, de coloribus infrascriptis, videlicet azurro de Alamania maioris rationis, auro fino in dyadematibus et stellis in tribuna et aliis locis, secundum voluntatem dictorum fratrum et ipsorum pictorum, in spatium duorum annorum proxime futurorum; on 6 November 1433 the syndic and friars of San Francesco made a final payment to Pietro di Angelo, who had bene et sollempniter painted the chapel ut ex picturis ipsius cappelle clare et evidenter apparet; for these notarial acts, see Felicetti 2001, p. 261-262, docs 40, 42.

66 In July 1434 Giovanni di Paolo di Maestro Nicola of Spello received 37 gold florins pro pictura tribune ecclesie Sancti Andree de Spello, see Felicetti 2001, p. 262, doc. 43.

67 On 22 May 1498, the painter Pierantonio de Mezzastris from Foligno agreed to paint the cappellam magnam of San Francesco, Bevagna with a Crucifixion at the top and 10 scenes from the life of Saint Francis below (including the Preaching to the Birds, doubtless because it was considered to be a local prodigy) and at the bottom frescium mediorum beatorum ordinis dicti ordinis, videlicet a medietate supradictorum beatorum, probably a frieze of illustrious Franciscans after the model of Gozzoli’s scheme at nearby Montefalco. The contract was drawn up in terra Mevanie in coro magno ecclesie Sancti Francisci. On 17 October 1499 Magister Perantonius Mezzastri de Fulgineo pictor cappelle magioris sive tribune altaris magioris ecclesie Sancti Francisci received partial payment and guaranteed the quality of the work for the next ten years. For full transcriptions, see Felicetti 2001, p. 274-276, doc. 83, 84.

68 For these later remodellings, see Carbonara 1982.

69 For the architecture of Sant’Andrea and the church’s later transformations, see Alici 1997.

70 For the church of Santa Maria and its frescoes, see Bianchi et al. 1994, p. 39-50.

71 See above, note 2.

72 This placement is suggested by the initial contract the Sansepolcro friars made with Antonio d’Anghiari on 9 October 1430, according to which the local painter agreed to paint the anterior face of the polyptych which overlooked the altar and faced the main door of the church: Magister Anthonius Iohannis de Anglario pictor habitator dicti Burgi promixit […] pingere et ornare […] tabulum positam ad altare maius dicte ecclesie Sancti Francisci, scilicet partem anteriorem que respicit altare predictum et portam maiorem dicte ecclesie. The document is transcribed in full by James R. Banker in Israëls 2009, II, p. 568-569.

73 For this document, see above note 17.

74 The so-called scripta only specified four Passion scenes for the front predella and four episodes from the life of Blessed Ranieri for the back, see Israëls 2009, II, p. 571-572.

75 The various options are considered by Israëls 2009, I, p. 171-173.

76 On 29 April 1578 the Sansepolcro friars commissioned the local woodworker Berto di Giovanni degli Alberti to fashion a ciboreum ligneum supraimponendum altari maiori. This notarial act, discovered and transcribed by Machtelt Israëls, is published in Israëls 2009, II, p. 576-577.

77 Cooper 2016, developing arguments advanced in Cooper (D.) 2005. For additional points in favour of a lost double-sided polyptych at Città di Castello and its likely dating, see De Marchi 2010.

78 The altar mensa at Città di Castello is 378 cm wide compared to 342.5 at Sansepolcro; the surviving narrative scenes in Chicago and London that can be associated with the altarpiece from Città di Castello are 62.9 cm wide compared to 52-53.5 cm for those from Sansepolcro, see Cooper (D.) 2005, p. 113, 120-121. A third scene that appeared at auction in Paris in 2007 is only 58 cm wide but has been cut down on all sides, see Cooper 2016, p. 107-108. Spinello’s altarpiece would have been replaced by a Eucharistic tabernacle, probably commissioned in connection with the re-consecration of the church in 1591; the tabernacle and the memoria of the 1591 consecration it contained were described c. 1726 by Alessandro Certini, Memorie delle chiese e monasteri tifernati, Città di Castello, Archivio Capitolare, MS 1, fol. 67v.

79 Solberg 1992 and 2018. See also the extensive entries by Vittoria Garibaldi (Garibaldi 2015, p. 289-296, cat. 100) and now Gail Solberg, Maria Cristina Tomassetti, Daniele Costantini and Kirsten Hinderer (Solberg 2020, p. 206-219, cat. 27).

80 Gordon 1982 and 2002, and now Garibaldi 2015, p. 128-134, cat. 5.

81 On the question of the predella’s central field, see Solberg 2018, p. 15-16, notes 29-30 (Solberg observes that the Perugia predella would have been too shallow to allow for a Eucharistic container or tabernacle). On the possibility of a central opening, see now Gardner Von Teuffel 2020, p. 104.

82 Frati 1886, p. 45-46: Ecclesia [San Francesco al Prato] hoc modo praeparata fuit, ut citra chorum apparatus fuisset Papae et Cardinalium; quoniam in choro Cardinales non bene sedissent; nec etiam commode, quoniam suggestus Papae super eo firmari non potuisset ob illius vetustatem, et fragilitatem. Tabula sive Anchona altaris propter plangorem fratrum moveri inde non potuit; fratrum maxima malignatas fuit. Suggestus fuit quatuor graduum in medio chori ante faciem altaris, sicut in sancto Petro Romae. Hinc inde sedilia pro Cardinalibus viginti, quoniam unus, videlicet sancti Severini, non adfuit; first cited in the literature on Taddeo’s altarpiece by Cooper 2001, p. 16, note 50.

83 For Raphael’s altarpieces in San Francesco al Prato, see Cooper 2010. The bequest for the high altarpiece appears in the will of Domina Panfilia filia olim Gentilis Francisci Nicholai de Perusia (4 July 1533), ASPg, Archivio Notarile 862, fol. 177v; see Cooper – Sartore 2020, p. 83. Unless it was combined with other bequests, the relatively modest gift of 25 florins would suggest a renewal of an existing structure rather than a commission for a new altarpiece.

84 For the contract of 2 October 1592 between the friars and “maestro Fiorenzo di Giuliano pittore perugino”, see Rossi 1873, p. 72-74, doc. 68 (mistaking the year for 1582); the opening clause (p. 73) required that Fiorenzo “facci il basamento con il piede di quatro parte et che si possa vedere il santissimo sacramento dal coro come è in disegno”. Rossi did not publish the clauses that preceded the contract, which asked Fiorenzo to “levar la tavola [dell’]altar grande” and to “segarla per mezzo in modo che le figure non siano lese et […] di sopra farvi li concicioni”; for further discussion, see Cooper – Sartore 2020, p. 84.

85 Gordon 1996; Garibaldi 2015, p. 192-197, cat. 32.

86 Gordon 1996, p. 35, citing Tabarelli 1977, p. 59, doc. XII: item reliquit quod in coro et super altare maioris fiat una tabula picta ex latere ubi dicitur missa aut ex utroque latere et alias prout videbitur guardiano dicti loci, pulcra et cum figuris, ut melius et convenientius videbitur.

87 The contract of 10 September 1502 was published by Canuti 1931, II, p. 237, doc. 379. For further discussion, see Gordon 1996, p. 35, and Cooper 2001, p. 18.

88 Canuti 1931, II, p. 237, doc. 379: “Item pengere in decta tavola, dall’altro lato verso la Chiesa delle donne, in mezzo di essa fare uno thruono cioè el Segnore et la Madonna, et da piej de essa tavola depengere 4 Apostoli et altre teste secondo remanerà el spazio, et da capo de essa tavola doj angeli et la cornice tutto similmente mettere ad oro, comme nella tavola dinanzi et in fine de detta tavola pengere doi figure et uno S. Bernardino, et l’altro uno Beato Bernardino de Feltrio, et in mezo el nome de Iesu [this seems to refer to a predella, which does not survive], et nel fine de essa tavola dove è el muro una pictura de color fine”.

89 Ibid.: “pingere tabulam existentem super altare magnum dictae Ecclesiae, in qua est Crucifixus et in ipsa tabula pignere quatro figure dal lato dove è el detto Crocefisso, cioè la figura della nostra Donna, S. Giovagne Battista [sic], S. Francesco et la Maddalena et appresso de le piaghe de le mano de detto Crocefisso doi angeletti, et la cornice de dicta tavola da capo, da piei, et da lato mettere ad oro, et el resto de dicta tavola pengere de colori fini et neri […]; da piej pengere doi angeletti, et pengere li sgabelli di essa tavola a la proporzione di esse figure.”

90 For the contract of 13 September 1470 between the Gualdo friars and Magister Nicolaus pictor egregius de Fulgineo, see Guerrieri 1930, p. 134-136, doc. 1.

91 Cooper 2001, p. 42-43. Documents cited by Guerrieri 1933, p. 373, indicate that the Gualdo house was raising funds for a new campanile concurrently with the high altar project. Further unpublished notarial acts suggest that the friars may have originally intended a smaller, less ambitious panel. In his testament rogated on 15 December 1463, Restorus Angeli Ser Martini de Gualdo […] reliquit ecclesie Sancti Francisci de Gualdo florenos duos expendendos per ipsos fideicommissariis propingenda una tabuletta parva pro altare maiori dicte ecclesie, et duos alios florenos pro pingenda alia figura beate Marie Virginis cum filio in brachio super et foris portam dicte ecclesie Sancti Francisci. Another bequest made on 24 April 1465 seems to look forward to Alunno’s commission: Domina Ludovica filia quondam Iacobi Francisci Grassi et uxor quondam Fiorii Grandis de Gualdo […] reliquit ecclesie Sancti Francisci de Gualdo florenos tres ad dictum rationem convertendos in pictura tabule que nunc depingitur. For these wills, see Archivio Storico di Gualdo Tadino, Archivio notarile, 2.1 (Ser Gaspare Ser Ranieri), fol. 151r-v, 159v. The quittance of 1471 published by Guerrieri 1930, p. 136 (doc. 2), records that Ser Gaspare’s mother, Fiordalisa, had given 10 florins in her own testament in et pro pictura tabule maioris altaris ecclesie Sancti Francisci di Gualdo. On the pawning of a chalice in 1472 to a certain “donna Ludovica” (probably the same as the 1465 testator noted above) in order to pay Alunno’s fee, see Angela Montironi’s entry in De Vecchi 2000, p. 107.

92 The contract specified “l’archicti che voglono sedere sopra lo scabello, alteza ped. I et mezo grassecto, computate base a pede”: Guerrieri 1930, p. 134. The significance of the open arcade for facilitating visual contact between the altar and the choir at Gualdo was noted by Christa Gardner von Teuffel (Gardner von Tauffel 2009, p. 227).

93 On these interventions, see Angela Montironi in De Vecchi 2000, p. 107.

94 It seems unlikely that the back was ever intended for public display. Montinori notes (but does not illustrate) “nella parte posteriore della predella sono visibili due tracce di scrittura, molto deteriorate, nelle quali è possibile riconoscere delle indicazioni di carattere pratico per l’assemblaggio del polittico” (De Vecchi 2000, p. 106).

95 See most recently Cannon 2009 and Neri Lusanna 2018. On the panels in the Ranghiasci collection, see Mambelli 2011, esp. p. 97-100. Observing the thinned state of the two saints now in the Worcester Art Museum and drawing on unpublished technical analysis and correspondence from the 2008-2010 collaborative research project on the altarpiece fragments funded by the Andrew Mellon Foundation and led by the Courtauld Institute of Art, Cannon 2009, p. 189, leaves open the possibility that the polyptych could originally have been double-sided. A rival candidate for the first single-sided Franciscan altarpiece in Umbria is the retable generally dated to the 1290s now in the Galleria Nazionale dell’Umbria with a nineteenth-century provenance from the Order’s convent at Farneto, see Garibaldi 2015, p. 164-168, cat. 19 and Delpriori 2015, p. 91-94. While the Farneto panel does not depict Saint Francis, the Passion cycle is closely related to the Maestro di San Francesco’s double-sided altarpiece for the Franciscans in nearby Perugia, see Gordon 2002, p. 231.

96 On the basis of the surviving predella panels, Cannon 2009, p. 189, allows that the altarpiece could have originaly been a heptaptych, which would suggest a setting over a high altar.

97 See the summary by Miklós Boskovits on the “The Franciscan Polyptych” in Boskovits 1984, p. 329: “[Carlo Volpe] rightly observes that the presence of S. Francis in the main register assures the provenance of these panels from a Franciscan church […] it follows that the painter executed his altarpiece for a Franciscan church in this region [Umbria], possibly at Gubbio or Assisi”. Cannon 2009, p. 190, has argued persuasively that the liturgical response implied by the text held by Saint Francis would have appealed to a Franciscan audience. The Franciscan provenance is also accepted by Polzer 2018, p. 49-52.

98 Mambelli 2011, p. 100, believed that the Ranghiasci connection “confirmerebbe infine l’ipotesi di Boskovits che vedeva la tavola essere stata eseguita […] per una chiesa francescana di Gubbio o dintorni”. For the legacy left to the Gubbio friars by Gabellus Ugolini Marcoli on 24 February 1345, see Rosati 1989, p. 63 (with an eighteenth-century summary reproduced as doc. 5,1) and now Neri Lusanna 2018, p. 92, and Cece 2018, p. 310 (with a transcription of the original document).

99 For San Francesco, Amelia, see Bozzoni 1985, p. 54-57 (the early dedication to Santi Filippo e Giacomo mentioned on p. 32).

100 See Canonici 1974, p. 281-282, for a local notarial act of 20 February 1324 concerning the foundation of a chapel in the Franciscan church at Amelia which refers back to an earlier document sigillatis sigillo conventus fratrum minorum de Amelia impresso, ipsis litteris cera rubea habente arculum sculpturas duarum ymaginum, videlicet sanctorum Jacobi et Philippi et in arculo has litteras: sigillum conventus fratrum minorum de Amelia.

101 Santi Filippo e Giacomo was under construction in 1275 and its dedication first appears in a bull of 1291, see Nessi 1962, p. 247-252.

102 For this large cross (380 x 265 cm) see Pisani 2016, with a suggested dating of 1295-1300.

103 On Fra Francesco, elected Provincial Minister for Umbria in 1316, see Nessi 1984a, p. 582-585. Chiara’s and Francesco’s father was called Damiano, offering a possible rationale for this saint’s inclusion in the altarpiece predella without his customary companion Cosmas. However, a significant weakness in the Montefalco hypothesis is the commissioning of a high altarpiece from the Maestro di Fossa for the new church in 1336, discussed below. The friars seem to have reused the monumental crucifix from their previous church, and one might assume they would have done the same with a high altarpiece if one was available, especially if they possessed a painting that may have been less than a decade old at that point.

104 Mambelli 2011, p. 99, believed that one of the missing saints on the predella should have been Saint Dominic, citing entry n° 370 from the 1877 Ranghiasci inventory: “San Lorenzo, San Damiano, Sant’Ubaldo e San Domenico: quattro tavole a tempera messe ad oro, alte centimetri 35 per 38, di Domenico di Cecco”. Discounting the attribution to the Gubbio painter Domenico di Cecco, the measurements match the surviving panels. It is difficult to know what weight to attach to the identification of the bishop saint as Gubbio’s patron Saint Ubaldus. If correct, this would indicate a provenance from Gubbio or nearby. As it happens, the only church in Umbria with a joint dedication to Saints Philip and James listed in the papal tithes imposed on the region in 1333 and 1334 was the hospital at Scheggia on the Via Flaminia, around 8 kilometres to the north-east of Gubbio, see Sella 1952, p. 183 (n° 2749), 196 (n° 2960), 201 (n° 3025), 215 (n° 3257). Mambelli, however, was unaware of a second bishop saint from a private collection, comparable in size, style and format to the other predella saints, that is now on long-term loan to the Courtauld Institute of Art, for which see Cannon 2009, p. 189, which would disprove the Saint Dominic hypothesis unless (as Cannon allows) the polyptych was originally significantly larger (reconstructed either as a heptaptych and/or as a double-sided altarpiece).

105 In addition to the comments by Cannon 2009, p. 189 (see above, note 96), Polzer 2018, p. 61, note 46, citing observations by Norman Muller, argued that the thinned state of the Worcester panels indicates that they were originally double-sided.

106 Nessi 1991, p. 101-117. Nessi (p. 108) suggested a connection with the consecration of San Francesco in 1340, with which Delpriori 2015, p. 251, concurred, noting stylistic similarities with a frescoed Crucifixion by the same artist at Campello sul Clitunno dated 1342. For the recent discovery of the donor inscription, see Cornini 2018, p. 83.

107 Reproduced by Todini 1989, I, p. 141 and II, p. 157, fig. 331 (as “già San Francisco, collezione Erland Sirèn… già Stoccolma, collezione Oswald Sirèn”). The picture, lot 31 with an attribution to Barnaba da Modena, was sold at Sotheby’s New York on 17 January 1985 for $18,700. The provenance was given as “Osvald Siren, and thence by descent in the family”. When Osvald Sirén acquired the piece is not known; he died in 1966 but had been publishing actively on early Italian painting from 1904.

108 Delpriori 2015, p. 248-253, noting the similarly tapered frames of the Sirén panel and the narrative scenes on the Vatican altarpiece. The Sotheby’s sale catalogue gives the panel’s dimensions as 13¾ by 12 inches or 35 by 30.5 cm. Allowing for the fact that the Sirén panel has been cut down, its dimensions bear comparison with the upper register of scenes on the altarpiece (just under 40 cm high and 42 cm wide). The Healing of John of Lerida is a rare scene in Franciscan vita cycles and is only found in extensive cycles (like the Upper Church at Assisi or San Francesco, Rieti). Depending on its central subject, a Franciscan scheme matching the Passion cycle on the Vatican altarpiece would require 9-10 scenes.

109 The presence of nails hammered through from the front of the retable (and then covered by gesso and paint) to fix transversal battons to the back removes any doubt on this point, see Cerreta – Baldelli – Alesi 2018. However, the likely existence of a matching retable with scenes from the life of Saint Francis, suggested by the Sirén fragment, still needs to be accounted for. One possible explanation may be that a double-sided altarpiece at Montefalco could have been composed from two structurally separate retables placed back-to-back, perhaps commissioned at different moments. I am grateful to Andrea De Marchi for ongoing discussion on this point.

110 See Nessi 2002, p. 57 (30 August 1511: pro acconcimine altaris magne), 62 (29 October 1516: pro ornamento altaris magne; 27 August 1517: pro pictura altaris magne), 63 (27 July 1518: pro pictura ancone altaris magne), 64-65 (20 December 1519: pro pictura tabule altaris maioris, quando pingetur dicta tabula; 2 January 1520: pro pictura tabule altaris magne; 17 February 1520: pro pictura tabula altaris magne), 66 (20 April 1520: pro pictura chone sive tabule cappelle magne; 2 August 1520: pro pictura tabule altaris magne, solvendos tamen quando pingetur dictum altare), 67 (27 April 1521: pro augumento pictura tabule altaris mayor; 15 May 1521: pro pictura cone in altare maioris existente in supradicta ecclesia Sancti Francisci), 71 (22 August 1526: pro pictura cone altaris maioris, et quod non possit aliquot modo divertere).

111 The contract of 30 October 1466 published by Nessi 1984b, p. 78 (doc. 13), stipulated: “Item che [the altarpiece] sia larga quanto che l’altare grande de Sancto Simone et altretanto alta, ciò tanto alta quanto larga da la banchetta in su conputando li colondelli da capo. Et [Vincioli] promette de fare etiam la decta banchetta […]. Item che in mezzo della decta tavola pegnerà la Vergene Maria col figluolo in collo con angeli dentorno. Et lu manto della Vergene Maria sia de brocchato d’oro fino o veramente de azzurro fino ultramarino fresciato intorno d’oro, con ragi et ornamenti d’oro per esso manto. Dal lato dritto san Pietro et san Francisco, dal lato sinistro san Paulo et san Ioambattista. In cima della tavola: in mezzo Dio patre et profeti, la Numptiata et angeli secondo li luochi dove meglio se aconvene. Dentorno et in la banchetta sancto Stefano, sancto Augustino et altre figurette.” The panel with Saints Paul and John the Baptist now in the Rocca Albornoziana was the left-hand side of the altarpiece’s main register; no other fragments have yet been identified.

112 Paoletti 1995.

113 For Piermatteo d’Amelia’s altarpiece, see Cassio 2005, p. 161-183. The contract for the painting composed on 29 September 1483 was published by Ricci 1987. The document prescribed “cinque fegure principali ad electa delli prescripti fratri et per li pilastri, cioè le colonne de fore, farci figurecte ad electa delli fratri et la predella della altare dipinta con figure ad electa delli fratri”; the work is dated 1485 by an inscription on the Virgin’s throne.

114 The drawings include a swirling vine on the back of the central panel and a male nude on the rear of the Saint Louis of Toulouse, for illustrations and discussion, see Cassio 2005, p. 166-167. The internal layout of the church at this time is unclear, a testamentary bequest of 1490 refers to an image of the Virgin que stat supra portam dicta ecclesie sancti Francisci unde intrant homines videlicet supra portam que stat iuxta pulpitum extra dictam ecclesia sancti Francisci which suggests that the nave was segregated by gender. The medieval cappella maggiore does not survive, in 1547 the communal authorities granted the lime necessary for the construction of the present apse, which is likely to be much deeper than the first. For these documents, see Cassio 2005, p. 322 (doc. XXI and XXIII).

115 For the contract of 16 April 1506 between the procurators of Sant’Andrea and Magistro Bernardino Benedicti alias Pintoricchio de Perusio, see Felicetti 2001, p. 276-277, doc. 89 (as well as the panel and the predella, Pinturicchio was asked to fashion unum tabernaculum pro reservatione Corporis Cristi in pedrella vel supra alibi melius; the altarpiece is still preserved in Sant’Andrea in the right transept). For Berto di Giovanni’s high altarpiece for San Francesco, Montone, now in the British Royal Collection at Hampton Court, see Henry 1996 (see p. 327-328, doc. Ia-Id for the documents recording the commission on 23 May 1506 and subsequent payments through to 15 March 1507).

116 For the stalls at Montone, see above, note 62.

117 See above, notes 72 and 86.

118 Israëls 2009, I, p. 243-253 (Polyptychs without painting: Sassetta, Piero della Francesca, and the rejection of unpainted altarpieces). As Israëls notes (p. 244), the carpentered structure Antonio d’Anghiari was asked to paint in 1430 may have been set over the high altar of San Francesco, Sansepolcro some four years earlier.

119 See Gordon 2002, esp. p. 239: “The reason for the commissioning of the double-sided altarpiece [for San Pietro] was almost certainly again emulation of the two local [Franciscan] altarpieces just discussed.”

120 See Gromotka’s contribution in this volume, fig. 18, for the reconstruction of the church interior around 1330 (where the presence of a tramezzo screen is included as a hypothetical possibility), also published in Gromotka 2015, p. 85, fig. 4 and in more detail in Gromotka 2018, p. 415-499. For an analysis of the apse and transept of San Pietro, see Cooper 2000, p. 209-212, where the pentagonal apse and transept are interpreted as an early fourteenth-century enlargement of the church, with corbelled vaults to facilitate a retrochoir. It formed part of the same programme of renewal that also included Meo da Siena’s altarpiece, a view echoed by Gromotka 2015, p. 111, note 13.

121 On the impact of the Congregration of Santa Giustina (later known as the Cassinese Congregration) on the choir arrangements at San Pietro, see Michael Gromotka’s contribution in this volume, and Gardner von Teuffel 2001, p. 124-126. The key pieces of evidence for Cassinese attitudes to choir arrangements are a late fifteenth-century plan for Santa Maria di Praglia showing a nave choir and a statement by a commission considering the rebuilding of Santa Giustina, Padua in January 1520, which noted that the high altar should be placed at the head of the church according to the custom of the congregation (altare maius statuatur in capite capellae majoris prout est consuetudo Congregationis nostrae), see Gardner von Teuffel 2001, p. 124, and for further detail, Winkelmes 2001; Bisson 2014b.

122 See Gromotka’s contribution in this volume (fig. 18); Gromotka 2015, p. 85-86, fig. 5 and in more detail in Gromotka 2018, p. 500-582.

123 Gromotka 2015, p. 80-94, esp. fig. 10 and Gromotka 2018, p. 587-635, superceding earlier reconstructions which interpreted Perugino’s altarpiece as a wall-like structure pierced by two lateral doorways to either side of the high altar mensa.

124 For the 1567 rearrangement, see Gromotka 2015, p. 94-98, fig. 14, and in more detail in Gromotka 2018, p. 690-707. The contract for the tabernacle agreed on 22 November 1567 with the Florentine woodworker Battista di Francesco was summarised by Rossi 1873, p. 65, doc. 58.

125 See Gromotka’s contribution in this volume (fig. 18), Gromotka 2015, p. 98-100, fig. 18, and Gromotka 2018, p. 690-731. The lavishly decorated choir stalls still in the apse were begun by the local woodworker Bernardino di Luca in 1525 and completed by the Zambelli workshop of intarsiatori from Bergamo in the 1530s, see Siciliani 2000 and Gromotka’s contribution in this volume (paragraph The sources). The contracts with Bernardino and the Zambelli were published by Rossi 1873, p. 44-46, 51-55, 59-60, doc. 28, 39, 44. As Rossi (p. 54-55) understood, these stalls were intended to be placed in the nave, before the high altar, and were only transposed to the apse in 1591.

126 For a comparison between the two, see Gardner von Teuffel 2004. For detailed accounts of Perugino’s Sant’Agostino commission and its slow progress, see Garibaldi 1997 and 2004.

127 The contract between the Augustinians and Mattia agreed on 23 April 1495 was published by Rossi 1873, p. 21-23, doc. 30. The terms of the document suggest that the ciborium was set at the summit of the structure and Mattia agreed “che la sopradicta Tavola sia lavorata tanto da uno canto quanto da laltro che non abbia niuna deformita”.

128 Cited by Garibaldi 1997, p. 120.

129 Rossi 1873, p. 26-28, doc. 1.

130 Agreed on 1 October 1502, excerpts transcribed by Rossi 1873, p. 27. The Perugian Augustinians appear to have had almost unlimited funds for their choir thanks to an unusual, uncapped bequest from Francesco di Cristoforo’s 1495 will, the generosity of which was disputed by Francesco’s heirs in 1507, see Mancini 2005, p. 96-97.

131 Rossi 1873, p. 27. An account of this protracted project is provided by Vitelli 2006, p. 231-233. The choir, probably largely executed in Florence, was finally completed in 1534.

132 Perugino renewed his original contract – from 1502 and now lost – with the Augustinian friars on 18 June 1512 and an outer casing or capssa was commissioned on 24 November that year. The polyptych seems to have been set over the altar by this date and the completion of its anterior face probably allowed mass to be celebrated. It is doubtful that the apse was usable, with the rear of the altarpiece unpainted and Baccio’s choir precinct unfinished, see Garibaldi 1997. On the capssa commission, see now Sartore 2014.

133 Bettarini – Barocchi 1966, III (1971), p. 607.

134 For the altarpiece’s later history, see Garibaldi 1997, p. 121-122.

135 Both Gurrieri 1957, and more recently Magrini 2014, p. 13, have argued that the spacious crossing and apse were fourteenth-century additions to a simple single-nave church completed by around 1300. The twin apsidal chapels at the heads of the transepts were frescoed at the end of the Trecento and their unusual arrangement finds a close analogy in the architecture of San Francesco, Ascoli Piceno, consecrated in 1371. In the most detailed study of the architecture to date, Sigismondi 2007, has drawn attention to the grant of 300 florins by the city’s magistrates for the completion of the apse in 1453 (p. 31, note 26). According to Pompeo Pellini (Pellini 1664, p. 573), in 1449 the priors “diedero dugento fiorini a’ frati di Sant’Agostino per far la tribuna all’altar maggiore, che essi havevano cominciata”.

136 Sulpizi 2014, p. 108-109: according to the concession, the Boncambi agreed quod dicti fratres possent quandocumque eius placuerit construi facere novam tribunam quiam inceptam post dictam cappellam libere et sine conditione; et in dicta cappella de Boncambiis costrui facere chorum et tota cappella et nova tribuna que quandocumque edificaretur sine intelligatur domus et de domo de Boncambiis. Giappesi (Sulpizi 2014, p. 105) also records that work on a new choir was foreseen by a significant bequest of grain in the testament by Francesco di Cristofano della Torre dated 5 January 1475 (“per fabricare vel rifabricare il nuovo choro” in the chronicle’s paraphrase of the source).

137 Sulpizi 2014, p. 108: the same document conceded pro tota domo de Boncambiis de Perusia usque ad finitam generationes cappellam magnam in qua est altare magnum cum dicto altare magno et pavimento ipsius terrestri ad usum dicte cappelle et pavimenti terrestris.

138 Banker 2001, showed how in 1451 the lay patron of the Augustinian altarpiece purchased the panel support that had been intended for a double-sided high altarpiece in San Francesco, Sansepolcro. The support had been commissioned in 1426, prepared for painting by Antonio d’Anghiari and Piero della Francesca in 1432, but then abandoned in 1437 (when Sassetta was contracted to make a new tabula identical to the first). In 1454 Piero agreed with the Augustinians to paint the anterior side of the panel only et non ex latere posteriori, a clause which can now be explained in light of Banker’s discovery that the wooden support had previously been prepared to receive images on both sides. The choir arrangements in Sant’Agostino, Sansepolcro are unclear, but Mazzalupi 2012, p. 7, 10, has recently published a notarial document of 1477 suggesting the presence of a choir screen in the church.

139 On the Pietralunga altarpiece, signed and dated by the artist, see Silvestrelli 1996. On Nelli’s cappella maggiore cycle at Gubbio, see Brenner 2007 (esp. the diagram reproduced as fig. 1 which shows the current choir stalls cutting into the picture fields of the lowest register). A single-sided pentaptych from the early fifteenth century attributed to the “Maestro di Narni del 1409” and now dispersed between the Vatican, Avignon and Munich can be linked with the high altar of Sant’Agostino, Narni, see Rossi 1994, p. 35-38 (cat. 9).

140 Delpriori 2015, p. 252-253.

141 The retrochoir was envisaged in Michelozzo’s initial plan for the Annunziata, begun in 1444, although its practical implications for lay access, conventual seclusion and liturgical observance would be much debated before its eventual completion in 1481 following revisions made by Leon Battista Alberti after the Gonzaga assumed control of the project in 1470, see Brown 1981 and 1996. For the 1400 contract for the Servite choir in Sansepolcro, see Israëls 2009, II, p. 588, doc. 10.

142 The reservations of the local Servites were expressed via their spokesman Giovanni Aldobrandini in a series of letters sent to the main patron of the tribuna, Ludovico Gonzaga, in early 1471. Aldobrandini argued that the rotunda design would create a series of difficulties: the radial chapels would be too small, the ambulatory was too narrow for easy lay access, laywomen would be in close proximity to the friars and not easily seen from the nave, the choir itself would be too small for the Servite community, and the entrance to the tribuna from the nave would be too narrow and have to be shared by the laity and the friars. This correspondence was published by Gaye 1839, I, p. 225-242. For detailed analysis, see Brown 1981, p. 98-108. In his letter of 23 March 1471 Aldobrandini submitted a revised plan which centred around a large cappella maggiore 24 braccia square to be allocated to the Gonzaga “dove si farà uno coro doppio, et ancora vi resterà spatio grande per li frati che stanno a lo leggio”; see Gaye 1839, I, p. 232-233, Brown 1981, p. 100. He does not specify where he would site the high altar but had earlier (p. 230) criticised its placement “alla entrata del coro […] benchè non sia suo luogo proprio” in the Michelozzo/Alberti scheme.

143 For these stalls, see Basta 2006. The relevant archival documents were summarised by Rossi 1873, p. 10, 14, 25-26, doc. 12, 22, 36. The detailed description of the church compiled by Fra Domenico di Francesco Baglioni in 1548 (Perugia, Biblioteca Augusta, MS 1232, fol. 63v) suggests the former presence of a tramezzo screen in the nave, see Frate 2006, p. 265. Duccio’s Virgin and Child with Angels from San Domenico, Perugia – painted c. 1305 and often thought to be the central section of a high altarpiece for the thirteenth-century church of San Domenico Vecchio – was part of a single-sided structure to judge from the traces of battens on the reverse, see Cannon 2013, p. 140-141.

144 For the Città di Castello polyptych, see De Marchi 1998, and Weppelmann 2011, p. 302-310 (cat. 77-78).

145 For Benedetto di Bindo’s high altarpiece for Santi Domenico e Giacomo, Bevagna, see Cannon 2013, p. 155-156. For Taddeo di Bartolo’s high altarpiece for San Domenico, Gubbio, see Solberg 2020, p. 258-261, cat. 42 (with reconstruction on p. 293).

146 Toscano 2008 and 2009.

147 Sensi 1988, esp. p. 192 for the construction of the main apse, begun c. 1435 and still ongoing in 1474, and p. 206-207 for an eighteenth-century account of the renewal of the choir and high altar prior to the reconsecration of the church in 1570. Further detail is provided by Marinelli 2015, p. 11-12, 55-57, clarifying that the choir was made in 1569-72 with later additions through to 1598, while the tabernacle was installed over the high altar in 1576.

148 For examples of polygonal gothic apses being added to Romanesque abbey churches, see Pirri 1960, p. 120-121; Gigliozzi 2011, p. 194-196.

149 The contract was published by Rossi 1874, p. 281-283. The Silvestrines commissioned the Marchigian master woodcarver Paolino di Giovanni of Ascoli and his assistant Giovanni da Montelparo facere construere et edifficare seu fabricare unum corum lignaminis in dicta ecclesia videlicet in tribuna dicte ecclesie videlicet circum circha dictam tribunam cum viginti septem sedibus magnis secumdum designum fattum in quodam foleo cartarum bonbicinarum factum per dictos magistros. Vitelli 2006, p. 220, is mistaken in claiming that these stalls were transported across Perugia with other Servite artworks and artefacts in 1542 from Santa Maria dei Servi shortly before its demolition to Santa Maria Nuova.

150 Rossi 1873, p. 5, doc. 6; the choir is signed and dated HOC OPUS FECIT PAULINUS DE ESCULO ET CUM SOTIO SUO IOHANE DE MONTELBERO TEMPORE FRATRIS COLE ANTONII DE PERUSCIO MCCCCLVI. Although remaining in their original context the stalls have undergone a number of alterations, see Vitelli 2006, p. 221-225.

151 Graham 1903, p. 38 (English translation), and p. 127, doc. 1 (original Latin text): et in altare magno dicte Ecclesie non est tabula que secundum condecentiam ipsius Ecclesie requireretur.

152 Graham 1903, p. 39 (English translation) and p. 128, doc. 2 (original Latin text): Item cum ad altare ecclesie S. M. Nove de Perusio porte Solis magnum sit quedem tabula non depicta, ibidem steterit per plures annos et propter dictorum fratrum, capituli et conventiis impossibilitatem et paupertatem dicti conventus non potuerit depingi.

153 The detailed iconographic guide attached to the contract of 9 December 1472 asked Fiorenzo to paint an Assumption and four other large figures “de uno lato” and a Virgin and Child and four saints “dall’altro lato”, published by Graham 1903, p. 41-44 (English translation) and p. 130-132, doc. 4 (original text in Latin and Italian). For the surviving elements of the altarpiece and summary of the commission, see Garibaldi 2015, p. 490-497 (cat. 185).

154 The fresh contract agreed between the Silvestrines and the artist on 8 August 1487 stipulated quam tabulam dictus Florentius teneatur et obligatus sit solum pingere ex uno latere; the document was published by Bombe 1912, p. 331-332.

155 On 27 September 1349 for the fee of 100 florins Niccolutius et Thomas Ciccoli Tardini de Assisio et porta S. Francisci et Angelus Cecce Polti de Assisio promiserunt et convenerunt […] facere et construere et ordinare in dicta ecclesia superiori (of San Rufino), videlicet in tribuna […] unum chorum de bono lignamine, omnibus eorum sumptibus et expensis, cum sediis illius amplitudinis et longitudinis ac etiam laborerio prout sunt facte sedie chori existentis in ecclesia superiori B. Francisci de Assisio, cum colundellis a parte inferiori ac etiam bactetoriis et cum duobus legeis parvulis. See Cenci 1974, I, p. 105.

156 For Alunno’s Nocera polyptych, see Cristina Galassi’s entry in Mancini 1996, p. 87-92 (cat. 38).

157 The situation for San Lorenzo is confused by Ottavio Lancellotti’s claim, made in the mid seventeenth century, that “il coro che stava prima nel mezzo della nave grande alla monastica, fu trasportato dove hoggi lo vediamo a spese del Capitolo e del Cardinale Francesco Armellini” (Armellini was made a cardinal in 1517 and died in 1527), see Lancellotti, Scorta Sagra, vol. II, Perugia, Biblioteca Augusta, MS 5, fol. 290v. Lancellotti’s testimony was discounted by first Rossi 1873, p. 20-21, doc. 29, and later Trionfi Honorati 1992, p. 281. Both scholars took a 1495 reference to the “serraglio del coro di San Lorenzo” as proof that the 1491 choir was always located in the cathedral apse, Trionfi Honorati noting that “i cori in abside sono rari ma non sono una eccezione prima del Concilio di Trento”. The retrochoir hypothesis has now been definitively confirmed by the close examination of the stalls undertaken by Fedeli 2015, p. 51-55: “si può dare per certo che il coro di Perugia sia stato concepito fin dalla sua origine per essere collocato nell’abside dove si trova tutt’ora.” According to Fedeli (p. 55) the “serraglio […] doveva essere una sorta di cancellata lignea intagliata, caratterizzata da quattro porte […] e doveva essere posto tangente alle due terminazioni del coro”.

158 The documentation presented by Mancini 2005 demonstrates that commissions for choir stalls in Perugia were carefully measured against other choirs within the city, at least in terms of ornament and pricing.

159 See Israëls 2009, II, p. 587-588, doc. 8, for the contract of 27 November 1400 for unum chuorum duplum et quadrum in ecclesia dicti loci de Sancto Angelo in Vado […] ad formam laborerii chuori novi facti in loco fratrum minorum de Burgo. It may be significant, however, that the contract for the Marchigian Servites did not specify the location of their choir, only that it should be double and square in plan. Another contract (p. 588, doc. 10) composed on 20 December 1400 between the Servites in Sansepolcro and the same family of woodworkers commissioned a set of choir stalls to be modelled on both the stalls in San Francesco and those planned for Sant’Angelo in Vado pro parte duplum et eo modo et forma quo et qua est chorus factus in ecclesia loci fratrum minorum de dicto Burgo post maius altare loci predicti et eo modo et forma quo et qua […] facere tenentur in loco fratrum Servorum Sancte Marie de Sancto Angelo in Vado (again the location of the Sant’Angelo choir is not specified, while the stalls in San Francesco are explicitly located behind that church’s high altar). See also above, notes 20 and 141.

160 For the Subiaco altarpiece, painted by Meo da Siena’s workshop around 1340, see Gordon 2002, p. 240-244. Gordon, p. 243, outlines several links between San Pietro and Subiaco which could explain the employment of the same Perugian workshop by both houses in quick succession.

161 Legislation on the location of choir stalls does not survive for the Franciscans at either general or provincial chapter level, but the later instance from 1501 of the Franciscan Observant provincial chapter for Emilia-Romagna permitting its house at Cotignola to retain its choir post altare maius prout nunc est on the condition that no other church should imitate it nisi de licentia et assensu capituli provintialis suggests that provincial authorities held competence in this area. For full discussion of the Cotignola case, see Cobianchi 2013, p. 26-27.

162 Graham 1903, p. 131, doc. 4; the significance of this passage was appreciated by Gardner von Teuffel 2009, p. 226-227.

163 In addition to the absence of surviving central predella panels from either Taddeo di Bartolo’s or Sassetta’s double-sided polyptychs, the 1502 contract for Perugino’s Monteripido altarpiece records that the predella facing the church was to be painted with the Observant saints Bernardino da Siena and Bernardino da Feltre to either side of the Name of Jesus, while the back was simply to have two angels with no specified subject at the centre, see Canuti 1931, II, p. 237, doc. 379, and above, notes 87-89.

164 Laschke 1992.

165 For Vasari’s double-sided Pieve high altarpiece, which incorporated a Eucharistic tabernacle at the centre of its predella accessible from both front and back, see Isermeyer 1950.

166 For a description of the altarpiece, see Laschke 1992, p. 203-208 and fig. 4 for Montorsoli’s preliminary plan for the altar and its pavement clearly showing the tabernacle recess at the back.

167 On lay access to conventual spaces as documented in notarial acts, see Israëls 2009, I, p. 129 (Painting for a Preacher: Sassetta and Bernardino da Siena); Cooper 2011, p. 100-104. Considering the degree of general access to the projected rotunda at Santissima Annunziata, Aldobrandini warned “non è cosa conveniente che le secolari donne con la presentia loro a’ frati sì propinque, et quasi con loro rinchiuse habbiano a dare loro alcuna turbatione”, while the laity that entered its narrow entrance “quasi si nascondono con li frati”, see Gaye 1839, p. 229-230.

168 Cooper – Robson 2013, p. 59-62. On the evidence pointing to the presence of a tramezzo screen in San Francesco al Prato, Perugia, see Cooper – Sartore 2020, p. 82.

169 Kempers 2016, p. 276-278, and also note 19 above. Kempers’s argument turns on the reading of the terms “cuoro grande” and “cuoro menore/piccolo” in the vernacular contract for the completion of the choir in San Francesco agreed on 24 June 1392, for which see Cooper – Banker 2009, p. 71, 587, doc. 6. The fact that the “cuoro grande” comprised 30 stalls with an additional “sedia grande” in the middle for a cappella maggiore approximately 10 metres square indicates that this term referred to a single bank of seating only, strongly implying that the term “cuoro menore” referred to the lower bank of seating in the same space.

170 Laschke 1992, p. 209-210.

171 See above, notes 31, 67 and 169.

172 For a valuable foray into this area, see Utari 2013.

Auteur

Donal Cooper

University of Cambridge - dac66@cam.ac.uk

Voir plus de livres

Le Thermalisme en Toscane à la fin du Moyen Âge

Les bains siennois de la fin du XIII^e siècle au début du XVI^e siècle

Didier Boisseuil

2002

Rome et la Révolution française

La théologie politique et la politique du Saint-Siège devant la Révolution française (1789-1799)

Gérard Pelletier

2004

Le juste marché

Le système annonaire romain aux XVI^e et XVII^e siècles

Monica Martinat

2004

Gilles Caillotin, pèlerin

Le retour de Rome d’un sergier rémois, 1724

Dominique Julia (dir.)

2006

Sainte-Marie-Majeure

Une basilique de Rome dans l’histoire de la ville et de son église (V^e-XIII^e siècle)

Victor Saxer

2001

Offices et papauté (XIV^e-XVII^e siècle)

Charges, hommes, destins

Armand Jamme et Olivier Poncet (dir.)

2005

La politique au naturel

Comportement des hommes politiques et représentations publiques en France et en Italie du XIX^e au XXI^e siècle

Fabrice D’Almeida

2007

Pittura rupestre medievale

Lazio e campania settentrionale (secoli VI-XIII)

Simone Piazza

2006

La Colonne Trajane et les Forums impériaux

Martin Galinier

2007

La Réforme en France et en Italie

Contacts, comparaisons et contrastes

Philip Benedict, Silvana Seidel Menchi et Alain Tallon (dir.)

2007

Pratiques sociales et politiques judiciaires dans les villes de l’Occident à la fin du Moyen Âge

Jacques Chiffoleau, Claude Gauvard et Andrea Zorzi (dir.)

2007

Souverain et pontife

Recherches prosopographiques sur la Curie Romaine à l’âge de la Restauration (1814-1846)

Philippe Bountry

2002

Le Thermalisme en Toscane à la fin du Moyen Âge

Les bains siennois de la fin du XIII^e siècle au début du XVI^e siècle

Didier Boisseuil

2002

Rome et la Révolution française

La théologie politique et la politique du Saint-Siège devant la Révolution française (1789-1799)

Gérard Pelletier

2004

Le juste marché

Le système annonaire romain aux XVI^e et XVII^e siècles

Monica Martinat

2004

Gilles Caillotin, pèlerin

Le retour de Rome d’un sergier rémois, 1724

Dominique Julia (dir.)

2006

Sainte-Marie-Majeure

Une basilique de Rome dans l’histoire de la ville et de son église (V^e-XIII^e siècle)

Victor Saxer

2001

Offices et papauté (XIV^e-XVII^e siècle)

Charges, hommes, destins

Armand Jamme et Olivier Poncet (dir.)

2005

La politique au naturel

Comportement des hommes politiques et représentations publiques en France et en Italie du XIX^e au XXI^e siècle

Fabrice D’Almeida

2007

Pittura rupestre medievale

Lazio e campania settentrionale (secoli VI-XIII)

Simone Piazza

2006

La Colonne Trajane et les Forums impériaux

Martin Galinier

2007

La Réforme en France et en Italie

Contacts, comparaisons et contrastes

Philip Benedict, Silvana Seidel Menchi et Alain Tallon (dir.)

2007

Pratiques sociales et politiques judiciaires dans les villes de l’Occident à la fin du Moyen Âge

Jacques Chiffoleau, Claude Gauvard et Andrea Zorzi (dir.)

2007

Souverain et pontife

Recherches prosopographiques sur la Curie Romaine à l’âge de la Restauration (1814-1846)

Philippe Bountry

2002