Siston is a small village in South Gloucestershire, located approximately 7 miles north-east of Bristol. St Anne’s church is located between Siston Court, an Elizabethan manor house, and the small cluster of cottages at the centre of the village. The standing church predominantly dates from the 13thc with additions and restorations from the 17th to 20thc. It comprises a chancel, aisleless nave, west tower, south nave porch and a south nave chapel, constructed largely of coursed stone rubble. The most significant 12thc survivals are the south nave doorway and the lead font.

The manor of Siston, located in the hundred of Pucklechurch, was held by a certain Anna prior to the Norman Conquest. By the time of Domesday Book (Moore, 1982: 168 a), it had been acquired by Roger I de Berkeley. As is common in Domesday, no church was recorded at this time. The manor later passed to Roger I’s son, Roger II de Berkeley (d. before 1130), and then to Roger II’s widow, Racendis. By 1130, the Berkeley honour had passed to William de Berkeley, nephew of Roger II, in the apparent absence of Roger II and Racendis’ son and heir, Roger III de Berkeley. On Racendis’ death, probably shortly before January 1138, the manor was acquired by Glastonbury Abbey through the agency of its abbot, Henry of Blois, also bishop of Winchester and brother of King Stephen, at a cost of 40 silver marks. Henry argued that Racendis had visited him personally and bequeathed Siston to the abbey on the event of her death, and the monks of Glastonbury also claimed Siston as an ‘ancient possession’. The transaction was subsequently confirmed by a royal charter issued in January 1138 (Cronne et al., 1968: no. 342, p. 130; Watkin, 1947: 129–30). In spite of this acquisition, Siston manor soon fell outside the control of Glastonbury Abbey due to the succession dispute between King Stephen and King Henry I’s daughter, Matilda. From 1139 and 1153, south Gloucestershire was effectively controlled by the Angevin party allied to Matilda and Siston manor appears to have reverted to Berkeley lordship. In 1153, Roger III de Berkeley assigned Siston as a jointure in the marriage-alliance between his son and the daughter of Robert fitz Harding. This transaction was challenged by the Glastonbury community and Duke Henry, the future King Henry II, ruled in favour of Glastonbury (November 1153). A compromise was subsequently reached in which the Berkeley family continued to hold the manor as a knight’s fee of Glastonbury Abbey. For a more detailed discussion of the complex dispute over Siston, see Stacy (1999) and Crouch (2000).

Alterations to the south nave doorway

Alterations have clearly been made to the south doorway, presumably when it was reset in the 13thc. This is revealed by an irregularity in the frontal chevron at the apex of the 2nd order, and the bare space around the winged tympanum. Presumably there was once a 1st order of voussoirs that have been lost. The inner section of the west impost (1st order) has clearly been renewed.

Iconography

The south doorway tympanum appears to depict the Tree of Life at its centre, a motif that is relatively common in the region and can be traced to Genesis (2:8—9) and Revelation (2:7, 22:2). The flanking crosses may symbolise the Crucifixion, although the skewed cross on the left-hand side could be an allusion to the martyrdom of St Andrew who was crucified on a crux decussata (X-shaped cross). Juxtaposing the Tree of Life with the Crucifixion may symbolise the redemptive power of Christ’s sacrifice. In this reading, the tree could also allude to Christ as the True Vine (John 15:1).

The figures on the font have been interpreted as apostles, their right hands raised in blessing and their left hands clutching books (Zarnecki, 1957: 11). Their identities are unclear. Christ in Majesty in often shown seated with hand raised in blessing, however these figures lack the cruciform nimbus and the mandorla, sometimes supported by angels, that typically characterise such a scene. It would also be highly irregular to present multiple depictions of Christ in Majesty on the same scheme. The beautifully carved arcades and scrolling foliage may have been designed to reflect the Kingdom of Heaven or New Jerusalem (Revelation, 3:12 and 21:2–21).

Relationship to other regional churches

Elements of the carved decoration at Siston suggest awareness of sculpture carved by the Dymock and Herefordshire Schools. The Tree of Life is found on several church tympana associated with the Dymock School, including Dymock, Kempley, Newnham-on-Severn (Gloucestershire), Rochford (Worcestershire), Yatton, Moccas (Herefordshire) and High Ercall (Shropshire), as well as the font at Bromyard, Herefordshire (Gethyn-Jones, 1979: 15n, 60, plates 19-21, 60-61). Tympana with tree motifs that have been associated with the Herefordshire School can be seen at Hereford Cathedral and Kilpeck (Gethyn-Jones, 1979: plate 18; Thurlby, 2013: 71-73). However, all of these trees stylistically differ from the Siston Tree of Life, tending to be more lush and complex in design. The closest parallel to the Siston motif in terms of style can be found on the tympanum at Upper Swell. It shows a bird eating fruit from a tree with five branches and is carved in the same recessed outline technique, although the tree lacks trilobed leaves and is more rudimentary than the Siston example.

Frontal chevron of a similar type to that on the Siston doorway can be seen at Kilpeck (S doorway and chancel arch) and on the Shobdon arches, which are attributed to the Herefordshire School. This type of chevron can also be traced to St Peter’s Abbey, Gloucester (now Gloucester Cathedral), where examples dating from c. 1122 can be seen in the nave and NW aisle. Upleadon church (Gloucestershire), which was dependent on St Peter’s Abbey, displays frontal chevron on the north nave doorway. That said, there are numerous examples of frontal chevron across the region, at English Bicknor, Leonard Stanley, South Cerney, Quenington, Condicote, Edgeworth, Withington, Rudford, Alstone, Stanley Pontlarge (Gloucestershire), Moccas, Bromyard (Herefordshire), Rock and Beckford (Worcestershire), to name several. One idiosyncratic feature found at both Siston and Kilpeck is the use of three-dimensional lateral chevron on the labels of the south doorway and chancel arch, respectively.

The profusion of saltire crosses at Siston can also be traced to the Dymock and Herefordshire Schools, although this is another common motif found across the region. A notable feature that is found at both Siston and Kilpeck is the application of saltire crosses to the upper register of chamfered imposts on the south nave doorways. Other examples of this feature can be seen at various Gloucestershire churches, including Postlip (S nave doorway and chancel arch), Turkdean (blocked N chancel doorway), and Alstone (doorway). Saltire crosses with beads are another notable feature of the Siston south doorway, and the same motif can be seen on the chancel arch at Kilpeck (1st order of arch). Again, there are other examples of this motif at Gloucestershire churches, including Moreton Valence (chancel arch stringcourse), Condicote (S nave doorway), Turkdean (blocked N chancel doorway), and Great Washbourne (S doorway).

Examples of cable moulding with beading motif, like that seen on the Siston doorway west capital, can be found at Upleadon (N doorway), Condicote (S doorway), and Beckford (chancel arch). The same Siston capital is triple scalloped with recessed shields like the west capital of the Upleadon doorway. Similar recesses can be seen on the outer capitals of the S nave doorway at Siddington church (Gloucestershire). It is also worth comparing the stylised recessed scallop capitals on the Siston font.

The Siston font is related to five other lead fonts in Gloucestershire, located at Frampton-on-Severn, Oxenhall, Tidenham, Sandhurst and Gloucester Cathedral (originally from Lancaut). All of the fonts were produced using the same block mould, however the Sandhurt and Lancaut fonts differ in arrangement. The Siston font is identical to those at Frampton, Oxenhall and Tidenham. In terms of style, comparisons have been made to (near-)contemporary illuminated manuscripts, such as the Winchcombe Psalter; pre-Conquest illuminated manuscripts, especially those associated with the Winchester School and Canterbury; and Mosan-influenced enamel work (Zarnecki, 1959: 11-12; Zarnecki et al., 1984: 86, 247). The practice of depicting human figures or apostles beneath arcades is fairly common on 12thc stone fonts across Gloucestershire and Herefordshire; for examples, see Rendcomb, Newnham-on-Severn (Gloucestershire), Hereford Cathedral and Orleton (Herefordshire).

Manufacture of the lead font

Zarnecki (1957: 4) broadly identified Romanesque lead sculpture as the work of goldsmiths. This view was challenged by Oman (1958: 103). While agreeing that the blocks, or dies, used to manufacture lead fonts were produced by goldsmiths, Oman argued that these dies were originally designed for the manufacture of other products, such as retables, shrines and metal book covers. It follows that old dies discarded by goldsmiths could be acquired by plumbers and put to other uses, namely the production of lead fonts. The use of old dies would explain various imperfections in the details of the font.

As stated above, the impressions for the Siston font mould were produced using the same die three times. The said mould could have been made from clay. Afterwards, the bottom was cast by turning the font upside-down, filling it with sand and pouring on hot lead (Zarnecki, 1957: 4, 10-11).

Dating and patronage

In terms of style, the Siston south nave doorway fits comfortably within a second-quarter 12thc context. The presence of frontal chevron, which seems to have been first used in the region at St Peter’s Abbey, Gloucester, c. 1122, indicates a date after c. 1125, while the simplicity of the tympanum and the profusion of saltire crosses, a motif which had broadly fallen out of use by the mid-12thc makes a date after c. 1150 unlikely. Unfortunately, there is no documentary evidence from this period to offer a firm date.

A member of the Berkeley family emerges as the most likely patron for the scheme since the manor belonged to the family from the later 11thc until 1153, with only a very short break between c. 1138 and c. 1139 when the manor was acquired by Glastonbury Abbey. A date in the first half of the proposed time frame would place Roger II de Berkeley and/or his wife Racendis as patrons. On the other hand, a date after c. 1138 would place William de Berkeley, their nephew, as the patron, while a date after the beginning of the 1140s would suggest Roger III de Berkeley who had emerged to claim the Berkeley honour by c. 1146.

A comparison with the major Berkeley foundation at nearby Leonard Stanley may prove fruitful here. The present author has attributed the earliest sculpture from the east end to the east nave, including the sculpted capitals in the chancel, to William de Berkeley, lord of the Berkeley honour from c. 1130 to the early 1140s and founder of the Cistercian abbey at Kingswood, Gloucestershire (Turnock). There is a distinct lack of chevron in the eastern arm and crossing of Leonard Stanley church, with frontal chevron first emerging on doorways at the western end which appear to date from the mid-12thc. On the other hand, the Nativity capital in the chancel exhibits stylised double scallop capitals with recessed shields that are comparable to the recessed triple scallop capitals at Siston. The concave beads that frame the Siston tympanum can also be tentatively compared to the convex varieties that decorate the west face of the east crossing arch at Leonard Stanley. Nonetheless, a closer relationship between the sculpture of Leonard Stanley and Siston might be reasonably expected if William was also the patron of the latter.

Racendis emerges as the likeliest candidate, although this identification is by no means definitive. Her reported dealings with Henry of Blois suggest a woman keen to safeguard her soul, while widowhood evidently presented her with new autonomy and resources that she could have directed into a small church-building campaign. Divergences between the sculpture of Siston and Leonard Stanley may have been a deliberate choice, especially if Stacy (1999: 12) is correct in detecting animosity between Racendis and her nephew, William de Berkeley. Instead, the motifs at Siston seem to have been drawn from a broad set of regional models, and the selection of frontal chevron and the Tree of Life motif suggests a desire to conform to regional trends of the 1130s, even though the sculptor(s) responsible lacked the skill possessed by leading members of the Dymock and Herefordshire Schools.

The issues associated with the Siston font make it impossible to judge whether the font was cast around the same time that the doorway was created. Zarnecki (1957: 32) originally dated the font between 1150 and 1175, however he appears to have revised this to c. 1130-1140 (Zarnecki et al., 1984: 247). Such a time bracket refers to the date at which the die was carved and not necessarily the date at which the font was cast. Theoretically, a considerable amount of time could have elapsed between the production of the die and the manufacture of the font if Oman’s theory is accepted. There is, however, no reason to challenge Zarnecki’s revised dating if it refers to the production of the die since the motifs and designs relate to sculpture and illuminated manuscripts produced in the region during the second quarter of the 12thc. There are also parallels between the font and the Siston doorway in terms of common motifs, namely saltire crosses, beading cable moulding and recessed scallop capitals.