St Radegund Abbey - A Re-Assessment of the Abbey Church

239 ST RADEGUND’S ABBEY – A RE-ASSESSMENT OF THE ABBEY CHURCH M. C. J. DAVIS In 1882 W. H. St John Hope carried out excavations on the site of St Radegund’s Abbey, near Dover, the results of which were published in Archaeologia Cantiana (St John Hope 1882). Fifty years later, this journal carried an article by S. E. Winbolt giving a description of the precinct and additional details which supplemented the work of St John Hope (Winbolt 1931). More recently, the Dover Archaeological Group examined the course of a pipeline that skirted the precinct (see Archaeologia Cantiana, LXXXVII) and in 2002 Keith Parfitt of the Canterbury Archaeological Trust carried out a watching brief during developments within the outer court. In neither case were there any significant monastic discoveries. In 2004, Parfitt and the Dover Archaeological Group conducted predevelopment investigations within the southern end of the east claustral range, the report of which will be published in due course. For the benefit of local charities, Mrs Grace Moynan of St Radegund’s Abbey Farm has produced a very useful pamphlet that enables visitors to make sense of the site and its history (Moynan 1995). Given the passage of time and the considerable increase in our understanding of monastic sites, a reassessment of the abbey’s remains is overdue. This article is devoted to re-assessing the evidence of the church, based on the author’s own observations and the expert advice from good authorities. St John Hope concentrated on establishing the ground plan of the claustral buildings with excavations of the chancel of the church, the demolished parts of the east range and the assumed site of the infirmary. Within three days he had ‘laid bare the walls of the eastern arm and the transept’. With similar alacrity he cleared the sites of the chapter house and the infirmary, then removed the rubble from the tower. Later in the same year the clearance of the church was completed, the finds included pieces of worked stone from main structures and internal features such as tombs. There is no account of where these worked stones were found and there is no record of the stones themselves. No archive appears to survive apart from some details in St John Hope’s notebooks held by the Society of Antiquaries of London (SAL MS 785/4). M.C.J. DAVIS 240 Fig. 1 St John Hope’s interpretation of the plan. ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 241 The Church St John Hope’s interpretative plan (Fig. 1) shows a monastic church that seems to have been put together in a haphazard fashion with little thought to style, logic or convenience. The plan shows a cruciform building with a short nave, transepts and a long solid-walled chancel containing both the high altar and Lady Chapel. Flanking this chancel are aisles with no obvious links to the main vessel. Each transept is graced with a single chapel. The tower, with curious flanking compartments on its eastern and western sides was inserted in the angle of the nave and the north transept. This plan gained sufficient acceptance to be included as an example of a church for canons regular in a general book on medieval monasteries (Cook 1961, p. 193). Careful examination of both the upstanding fabric and St John Hope’s plan will highlight anomalies that point to at least one very different interpretation. Unfortunately, most of the primary evidence, or what survives of it, is back under the turf and cannot be used at present to support the interpretation offered here. Where it has been possible to do so, the St John Hope plan has been shown to be accurate, so we can use it with reasonable confidence for assessing those parts that cannot be seen without re-excavation. The North Transept Recent clearance of ivy from the east face of the surviving part of the north transept wall (Plate I) has enabled a paper reconstruction of part of it. This advantage may not have been available to St John Hope. Due to the delicate nature of the fabric the survey was carried out avoiding contact as far as possible. Vertical off-sets from a notional base-line were employed, with a combination of measuring tape and plumb line attached to a bamboo pole, and the farm’s fork-lift. This was a less than satisfactory method because the long plumb line (used for horizontal measures) was affected by the slightest breeze and we were not able to see the top parts at eye level. Nor was it possible to establish the precise floor level without excavation. However, we were able establish the approximate height of the building and to create a workable elevation of a single bay (Fig. 2). The walls are of rubble with Caen stone string courses and dressings around the window. The much-damaged lower corbel supporting the wall-shaft under the vaulting springer and the surviving dressings for the western respond of the north crossing arch are of Kentish Rag. It is very clear that the church was intended to be vaulted and fenestrated in the elegantly simple style that characterised many monastic churches of the late twelfth century onwards, especially those within, or influenced by, the Cistercian tradition. The north transept fabric, however, is quite complex. Within the M.C.J. DAVIS 242 West wall of the north transept PLATE I ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 243 Fig. 2 Profile, north transept bay. M.C.J. DAVIS 244 surviving bay is a slightly off-centre arch that gives access to the tower. To the immediate south of it is the outline of what looks like an earlier arch, of which no voussoirs survive. Their absence is interesting. It would have been a brave act to remove voussoirs before blocking – normally they would have been left in situ and the whole area cosmetically rendered over after blocking. It is possible that economy demanded the re-use of the voussoirs. Alternatively, the feature was not an arch but a part of a simple wall arcade that was filled in and from which the voussoirs were easily removable. This feature is discussed further in connection with the nave. Howard Jones pointed out that the nature of the wall surfaces above and below the lower string course are different. The flint work of the upper parts is much finer than that lower down. This suggests either a long break during construction, or that the transept was subject to a major remodelling. The moulding of the exterior reveal of the large surviving lancet has been dated to between 1220-1260 (Dr Richard Morris, pers. comm.), which would indicate thirteenth-century remodelling of a Phase 1 church that had been constructed within perhaps a decade of the abbey’s foundation in 1191. This remodelling becomes more obvious when we consider the chancel, but a point to note here is that the east-west axis through the central vaulting springer in the north transept does not line up with the interior wall of the north transept chapel. The Chancel Much of the evidence supporting a reinterpretation lies in St John Hope’s plan (Fig. 1). Looking at each outer presbytery aisle wall it will be seen that the part dividing the aisle from the transept chapel is narrower than the rest of the wall. This suggests either inept mason-craft or a multiphased wall. It is argued that the aisle walls running east from the transept chapels were built later, suggesting an extended presbytery. The ‘haphazardness’ of the ground plan is further challenged by the evidence of simple geometry. The classic √2 system is apparent in the basic layout of the nave, transepts and the east range (cf., Fernie 1990). The diagonal across the cloister, when swung north through 45º, marks out the line of the inside face of the nave north wall. A similar process marks the inside face of the outer wall of the east range (the layouts of the south and west ranges raise complex issues concerning dating and phasing). This was a common practice, even among the humblest of monastic sites (Stalley 1990). In the case of St Radegund’s, the √2 logic based on the cloister and crossing diagonals does not seem to have been extended to the chancel. This is further, albeit negative, evidence that the chancel was built to a different geometrical scheme sometime after the crossing, transepts and nave. The length of the surviving west bay ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 245 in the north transept is approximately 5m. Each of the six chancel bays are, according to St John Hope’s measurements, approximately 4.4m long. If the length of the north transept bay was intended as the standard throughout the church then there can be little doubt that the chancel was constructed as a separate phase, replacing an earlier, shorter chancel. This was a very common practice in English church-building tradition. St John Hope’s proposal of solid walls in the chancel instead of arcades must be challenged. He drew what he found and his ‘solid walls’ showed up very clearly as parch-marks in the summer of 2003. But he missed a vital piece of negative evidence. He could not find any obvious entrance to what he described as the Lady Chapel, therefore he concluded that there ‘must have been doors’ leading from the side-aisles. He failed to note that he would have been excavating below the level of these doors (had they existed), therefore below the level of the floor itself. More likely, therefore, the ‘solid walls’ were sleeper walls upon which the chancel arcades were built – a very common practice to aid stability. If this was the case then the chancel starts to make better sense, leaving only concerns over the form of the original chancel and the cramped liturgical arrangements proposed by St John Hope. For their own churches canons regular frequently adopted the Cistercian ‘Bernardine’ layout. Such a plan is proposed for the Phase I church at St Radegund’s (Fig. 3). The narrow inside walls of the transept chapels mentioned above would have been the dividing walls of paired chapels. The surviving engaged column that once supported the northern end of the crossing arch into the nave rests on a corbel about 2.5m above the floor level. This suggests that the choir stalls originally intruded into eastern part of the nave – the usual practice in Bernardine churches (St John Hope placed the stalls within the crossing only). The geometric relationships between Bernardine chancels and the rest of the churches to which they belong vary enormously, so there is no simple rule that we can use to locate the end wall. However there is a near-match in the geometrical relationships between the chancel and transepts of St Radegund’s and those of Bordesley Abbey, Worcs. (Hirst et al. 1983, p. 228) assuming the end wall position is the same as that of St John Hope’s high altar ‘reredos’. In both cases, an angle of 100º is formed by the lines between the outer western corners of the transepts and the centre of the inside face of the chancel east wall. Comparing two churches a considerable distance apart may be stretching analogies a bit far but there is a case to be made for St John Hope’s high altar to be the first high altar and the ‘reredos’ to be a surviving fragment of the first end wall. The small gaps between the ‘reredos’ and the side walls (Fig. 4) may have been simple robbing of parts of the east wall footings when the foundations of the new sleeper walls were laid. Both of these surviving wall footings and the altar platform could have been buried under a new M.C.J. DAVIS 246 Fig. 3 Suggested outline plan, Phase 1 church. ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 247 floor laid for the extended chancel. If the chancel floor was raised then there is an even stronger case for sleeper walls under the chancel arcades, particularly if the aisle floors were set slightly lower. Unfortunately, St John Hope’s clearance of the chancel site may have destroyed critically important floor level evidence. No evidence that levels were recorded during the original excavation has come to light. Chancel extensions not only created greater and more flexible liturgical spaces but also allowed the community to move the choir stalls into the chancel, freeing up the crossing space. This could have been the case at St Radegund’s; with the new high altar being that of St John Hope’s ‘Lady Chapel’, behind which was a substantial free-standing reredos and a space for a pair of minor altars (Fig. 5). The location of the Lady Chapel, if there ever was one, cannot be ascertained on current evidence. St John Hope (p. 148) mentions, almost in passing, the foundations of very large rectangular footings to carry flying buttresses to support the extended east end. His reason, ‘…to carry the thrust of the roof’, seems a little inadequate. Buttressing of that magnitude suggest either a tall building or heavy internal vaulting, indeed both. Leland’s observation (quoted by St John Hope, 142), ‘…the quier of the chyrche is large and fayr …’ lends support to a tall, vaulted structure, the height of which can be judged from the surviving part of the north transept, if these two elements were constructed as part of an integrated scheme. There is no precise dating evidence for the construction of the new chancel apart from the clasping buttresses supporting the east end which would have gone out of fashion in the early thirteenth century (T. Tatton-Brown, pers. comm.). Fig. 4 Part of chancel from St John Hope’s plan, showing gaps between high altar ‘reredos’ and side walls. M.C.J. DAVIS 248 Fig. 5 Proposed Phase 2, extended chancel. ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 249 The evidence of the east end, the window in the north transept and the mismatch between the north transept vaulting and the dividing wall of the original pair of north transept chapels suggest that the initial ‘Bernadine’ east end was replaced with an extended presbytery in the early thirteenth century, the scheme being rounded off with remodelled and matching transepts some time after 1220. There is, however, a problem with the Kentish Rag dressings. The received wisdom is that the Rag was not used for fine dressings before the beginning of the fourteenth century, although it was used for weatherings and quoins in Canterbury during the thirteenth century (Tatton-Brown 1990, 77). This is at odds with the moulding evidence of the surviving north transept window. There are at least three possible resolutions to this conundrum: the transept was remodelled 100 years later with very archaic features; or the vaulting and the new crossing arch were installed 100 years after the transept was remodelled; or we need to rethink the dating of the first appearance of Kentish Rag in finely moulded form. The use of seriously out-of-fashion moulding design seems unlikely and the transept wall surface shows that vaulting was certainly intended when the present upper parts were constructed. Therefore, we may need to challenge the accepted view of the earliest dates for moulded Kentish Rag. The Nave The nave seems short compared with those of other Premonstratensian churches (e.g. Bayham, Shap and Torre). The thickness of the walls on the north and west sides are 10cm less than in that of the south side and the transepts. The presence of a piece of dog-tooth moulding as core rubble in the west wall suggests a rebuilding of an earlier nave, unless it was part of post-medieval blocking. The height of the nave wall on the north side against the tower seems to be about 3m lower than the north transept (Plate II). This may be the height of the main walls of the first church but closer survey of the top of this wall is needed before we can be certain. Unless it is hidden under the heavy coating of ivy there is no evidence of any bay profiles in the north nave wall, either belonging to the original church or anything similar to that surviving in the north transept. On the external west face of the north transept there is a shallow inclined line of scarring along the face of the wall that cuts across the lower part of the window. It was for a shallow-pitched roof of a ‘leanto’ structure against the north side of the nave, possibly along the west side of the transept also. The blocked feature in the transept discussed above could have been the access to this feature. A resistivity survey (McNaughton et al. 2003) indicates linear anomalies parallel to the north nave wall that might be evidence of an earlier aisle that was removed when the tower was built. M.C.J. DAVIS 250 The west wall of the nave is well bonded to the south wall, presenting a homogenous structure. This fact argues against any idea that the nave was once longer, but the abnormal shortness of the present nave and a strong but indeterminate anomaly highlighted in the McNaughton survey dissuades one from discounting the possibility. The extra thickness in the walls in the north-east corner of the west range suggest a break in construction, raising the possibility that a longer nave was intended for the first church but never completed. However, as the west range is likely to have been the last of the claustral buildings to be completed in their final form, these differences in wall thickness within the west range could have been simply a matter of choice by the later builder. Bishop Richard Redman (d.1505), Abbot of Shap and commissary-general of English Premonstratensian monasteries, commented unfavourably about the dilapidated state of the buildings (Gribbin, 2001, pp. 91, 129) and Leland, some years later, observed that the buildings, though in good condition when he saw them, had been ‘more ample than they be now’. He does not elaborate but he could have been referring to the church. The Tower The insubstantial nature of the surviving remains of the north-west crossing support argues against the likelihood of a central tower, even a Tower, south side PLATE II ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 251 low lantern. In common with many canons’ regular churches, towers were erected elsewhere, normally at the west end (cf. Beauchief, Shap, Torre, Ulverscroft, Lilleshall, et al.). Blanchland’s tower is set against the end of the north transept, so the placing of St Radegund’s tower should not be regarded as particularly eccentric (St John Hope quotes a number of Kentish parish churches whose towers are in a similar position). The two narrow wings on the east and west sides are unusual and more research is needed to make sense of them. Any external access has been destroyed by the construction of post-medieval brick arches. Internal access is by one arch into the north transept and another into the nave from the western flanking wing. This latter arch is curious; it has a hollow-roll hood mould of Caen stone with an under-built order of Kentish Rag with tile-creasing separating the two. Its present form and location conforms to the overall tower structure, giving a strong hint that the hood mould at least might have been recycled from an earlier feature. If this was the case, and taking into account the absence of any evidence of north wall windows, then the eastern half of the north nave wall, if not the whole of it, was rebuilt when the tower was constructed. Three further points are worth noting. First, the tower and its wings provide a through-route from the nave to the north transept, possibly bypassing a blocking of some kind at the east end of the nave. Second, from the evidence of the blocking material and the robbing of worked stone, the surviving north transept window was probably blocked up after the monastery’s closure. Third, as the ground floor of the tower was vaulted the only access to the upper floors is via an opening in the north nave wall at first floor level, interpreted by St John Hope (p. 146) as the access to a pulpitum. The presence of both a pulpitum and a choir screen in a church of this size must be questioned, as must St John Hope’s assertion that the Gospel was read from such a feature in Premonstratensian churches. This first-floor access to the tower begs a question about the use of the nave in the latter part of the abbey’s life, particularly as the community at this stage was rarely more than 10-strong (Gribbin 2001 p. 51). Conclusions The uncritical acceptance of St John Hope’s conclusions led to a suggested reconstruction reproduced in Moynan (1995). This reconstruction, though true to the final ground plan, fails to portray the true nature of the building or its constructional history. Based on the evidence and arguments given here, an axonometric interpretation of the church’s constructional sequence is shown in Figs 6-8). The outline sequence proposed is as follows: Phase 1 (c.1191-1200): simple ‘Bernardine’ church with a short aisle-less nave. M.C.J. DAVIS 252 Fig. 6 Axonometric projection, Phase 1. ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 253 Fig. 7 Axonometric projection, Phase 2. M.C.J. DAVIS 254 Fig. 8 Axonometric projection, final Phase. ST RADEGUND’S ABBEY - A RE-ASSESSMENT OF THE ABBEY CHURCH 255 Phase 2 (early thirteenth century): replacement of the original chancel and the inner transept chapels with a 6-bay chancel with aisles covering four bays, and the remodelling of the upper parts of the transepts to match, increasing the height by up to c.3m. The whole new work was vaulted throughout. Phase 3 (date unknown): additional structure built on the north side of the nave and/or north transept, possibly as an arcaded aisle. Phase 4 (? fifteenth-century): construction of the tower and possible major reconstruction of the north and west sides of the nave. At some stage, heavy flying buttresses, as St John Hope has pointed out, were erected to prop up the eastern half of the chancel. Much of the detail is speculative but there is enough evidence to support the basic shapes and layouts given in the illustrations. The gabled weathering for some of the thirteenth-century buttressing is suggested from the evidence of the surviving buttress against the north transept (visible only from ground level within the confines of the tower). The temptation to include a flèche over the crossing was resisted. The flattened roofs in the final phase reflects the likely problems faced by a church that was feeling its age and had been subject to neglect. The suggestion of the Kentish ‘firebox’ tower seems appropriate, given the monastery’s cultural environment. The survey techniques supporting this paper were admittedly crude but the hazards within the surviving structures were a major inhibiting factor. Much more study needs to be done, both above and below ground, to understand the nature of this seriously misunderstood church. Given its fragile state a detailed survey of the tower is essential. A strong effort to hunt down and record the missing architectural fragments found by St John Hope may pay archaeological dividends. A re-excavation of at least critical parts of the church and the area to the north of the nave should answer many questions suggested here. From the evidence so far and given the high-ground landscape in which it was set, we can conclude with some confidence that the church of St Radegund’s Abbey had been a very significant and even dramatic architectural statement, and one of the largest and most impressive rural churches in east Kent. Its status as such needs to be recognised and acknowledged. ACKNOWLEDGEMENTS The author has a number of people to thank: Peter Kendall (English Heritage) and Keith Parfitt (Canterbury Archaeological Trust), Howard Jones, Tim Tatton-Brown and Richard Morris for their valuable advice and encouragement; Susannah Davis, Richard Davis and the late Virginia Huntley for their invaluable assistance during the survey; Brian M.C.J. DAVIS 256 and Carol McNaughton and Mick Webb for the very useful geophysics support. St John Hope’s account of his excavation will raise concerns among today’s archaeologists and one or two of Winbolt’s speculations are highly questionable. However, in their days, both these men were pioneers blessed with great enthusiasm for, and love of, the subject. We are still learning how to improve archaeological investigations, so this article is written with gratitude for what they achieved. Above all, the author is hugely grateful to Albert and Grace Moynan of St Radegund’s Abbey Farm for ready access to their property, their kindly forbearance and for their very generous assistance and hospitality. REFERENCES Cook, G.H., 1961, English Monasteries in the Middle Ages, Phoenix. Fernie, E., 1990, ‘A Beginner’s Guide to the Study of Architectural Proportions and Systems of Length’, in Fernie and Crossley, 229-237. Fernie, E. and Crossley, P., 1990, Medieval Architecture and its Intellectual Context, The Hambledon Press. Gribbin, J.A., 2001, The Premonstratensian Order in Late Medieval England, Boydell. Hirst, S.M. et al., 1983, Bordesley Abbey II, BAR British Series 111. McNaughton, C. and B., and Webb, M., ‘St. Radegund’s Abbey, Resistivity Survey’, unpublished report for English Heritage’, dated 8 December 2003. Moynan, G., 1995, St. Radigund’s Abbey, Dover, Kent (Premonstratensian Order), privately printed. St J. Hope, W.H., 1882, ‘On the Praemonstratensian Abbey of St. Radegund, Bradsole in Polton, near Dover’, Archaeologia Cantiana, XIV, 140-152. Stalley, R., 1990, ‘Gaelic Friars and Gothic Design’, in Fernie and Crossley, 191-202. Tatton-Brown, T.W.T., 1990, ‘Building Stone in Canterbury c.1070-1525’, in Parsons, D., Stone Quarrying and Building in England AD 43-1525, Phillimore, 70-82. Winbolt, S.E., 1931, ‘St. Radegund’s Abbey, Dover’, Archaeologia Cantiana, XLIII, 187-198.