Second cathedral is Early Medieval

 

Summary. Building a heating system beneath the choir floor in 1854 revealed the foundations of a large apsidal building whose dimensions, proportions, construction methods, and underlying metric system strongly indicate an Early Medieval date. The layout conforms to the short perch (15‑ft) grid used in Early Medieval planning between c. 600–1020. Its proportions closely resemble Brixworth Church, dated late 8th or early 9th century. Metric, structural, and comparative architectural evidence makes a Norman date untenable and instead gives a construction around c. 770, plausibly during the reign of King Offa.

 

The installation of an underfloor heating system in 1854 led to the discovery of early church foundations below the choir floor. A massive choir- apse foundation was described by the stonemason John Hamlet, with assistance given by James Rawson a local physician and George Clarke, the clerk of works to George Gilbert Scott the architect for the work.

Hamlet drawing with dimensions[1] 

Hamlet drawing superimposed on the layout of the current cathedral

In August 1859, the architectural historian Robert Willis examined the remains, took measurements and published his findings in 1861.[2] Willis saw only the south‑east portion of the apse, whereas Hamlet and Rawson had observed more extensive sections in 1856 and 1860.

Robert Willis  

Willis’s 1861 drawing of the choir-apse foundation with dimensions. The foundation was internally 15.9 m (52 ft 3 in.) wide and 21.4 m (70 ft 1.5 in.) long. Both the north and south walls measured 1.7 m (5 ft 8.5 in.) wide which gave a total span for the choir-apse of 19.4 m (63 ft 8 in.). Willis measured the foundation height as c. 1.8 m (6 ft). The black parts of the pillars were dated to be ‘Early English’ and the pillars for the foundation church were either timber and now lost or stone and incorporated within the Early English pillars.

 

The outer edges of the lateral walls appeared to align closely with the bench tables of the present side aisles, with an estimated gap of c. 7 in.[3] The apse was perfectly semicircular, matching the width of the straight choir walls—a basilical form unparalleled among surviving British cathedral foundations but consistent with 8th–9th‑century Early Medieval churches.[4] The foundation was described as: buff‑cream, dense, concrete‑like mortar, that was extremely hard[5]; consisting of rubble and mortar poured into a trench;[6] with upper levels faced with rough ashlar over a poured concrete core; and devoid of dressed, uniform stone typical of post‑10th‑century Norman work. This is extraordinary.

 Photograph taken 1994 of the foundation in the north choir aisle.

 

This construction has similarities with the early 7th‑century apse at Rochester,[7] the 9th‑century crypt at Wing, built of “rudest rubble masonry,”[8] the foundations at Deerhurst, containing 5 ft or more of coursed rubble, and the rubble‑built walls of Brixworth.[9]  The absence of Norman masonry[10] characteristics is striking.

AI rendition of how the second cathedral looked

 

 

The Short Perch Grid: A Diagnostic Early Medieval Metric

Archaeological work since the 1970s, including  at Catholme, 8.5 km from Lichfield, has demonstrated that Early Medieval buildings, c. 600–1020, frequently used a short perch of 15 ft (4.57 m) as a measuring module.[11] Thus, Catholme’s, c. 680–700, buildings align to a 4.57 m grid; a 15‑ft grid appears in major churches across Kent, Northumbria, Mercia, and East Anglia between 600–800; the metric reappears 950–1020 in manorial and settlement planning and it disappears entirely thereafter. By contrast, the long perch (18 ft / 5.48 m) was introduced by the Normans in the 12th century.[12]

Catholme showing alignment to a short perch grid. Thanks to J. Blair[13].  

St Pancras, Canterbury, 7th-century showing its dimensions on a short perch grid and the remains of the early church.[14]

At least five short‑perch measurements are identifiable in the foundation.

1.     Choir–apse length

2.     Internal choir width

3.     Cross‑distance between drum columns

4.     Cross‑distance between abutments on the choir west wall

5.     Distance from choir west wall to apse chord

Crucially, no long‑perch, 18‑ft, measurements appear anywhere in the foundation. The 75 ft 1.5 in. distance from the apse exterior to the choir west wall equals 5 short perches. The 52 ft 3 in. internal apse width is 3.5 short perches minus 3 in. The choir fits a 12‑perch grid, the apse an 8‑perch grid. This is decisive evidence for an Early Medieval layout.

Apse with short perch gridding. The grid is drawn with 15 feet squares.

 Wall Thickness and Structural Reconstruction

No walling was recorded in 1854–61, so its thickness must be inferred.[15] Early Medieval walls are typically 2 ft 6 in.–3 ft, but exceptional examples reach 3 ft 9 in.–3 ft 11 in., including: Deerhurst apse (3 ft 9 in.), South Elmham nave (3 ft 10 in.), Brixworth (3 ft 4 in.–3 ft 10 in.), Northampton palace (3 ft 11 in.), and Wenlock Priory (nearly 4 ft). The best fit for Lichfield is 3 ft 11 in., matching the wall of the rectangular chamber abutting the apse. This suggests: foundation width = 1/3 short perch and wall width = 1/4 short perch.[16] This proportionality is unlikely to be accidental.

Best fit short perch dimensions for a possible wall.

Comparison with Brixworth Church

Rodwell argued pre‑Conquest churches often shared master‑plans across regions.[17] Brixworth, one of the largest surviving Early Medieval churches, aligns closely with the short‑perch module[18] and shares numerous proportional similarities with Lichfield:

Feature	Brixworth	Lichfield Second Cathedral
Overall width	c. 63 ft	c. 63 ft
Nave/inner apse span	30 ft (2 short perches)	30 ft (2 short perches)
Choir/presbytery bay	30 ft square	30 ft square
Apse + choir length	c. 69 ft	75 ft
Ambulatory width	7 ft 6 in.	6–7 ft
Wall thickness	c. 3 ft 9–10 in.	c. 3 ft 11 in.
Rubble construction	Yes	Yes

The distance between the inside edge of the piers at Brixworth is the same as the estimated distance of the assumed piers inside the current nave columns at Lichfield.[19]  It is thought at Brixworth there was a transverse wall with triple openings (a larger central choir-arch and two lateral door-size openings) separating the nave from the choir bay. This arrangement might have occurred at Lichfield.[20] Distance between the nave and choir columns at Lichfield is almost 75 ft or 5 short perches. It is not very different from Brixworth’s 22.8 m (75 ft) length for the choir-apse. Plaster was evident at Brixworth and Willis thought it could have been added externally to the apse at Lichfield.[21] Brixworth’s date, late 8th or early 9th century, was fixed by radiocarbon dating of charcoal in mortar.[22] There are more parallels and too numerous to be coincidental.

Brixworth and its short perch grid. Thanks to J. Blair¹⁴.

Brixworth contrasted with Lichfield.

Historical Implications: Offa and the Mercian Basilica Tradition

If Brixworth, late 8th–early 9th century, and Lichfield share a master‑plan, then Lichfield’s second cathedral likely belongs to the same Mercian basilica tradition, which includes: Brixworth, Cirencester, Deerhurst, and Wing,[23] These churches have walls over 3 ft thick, often include apses c. 4 ft thick, exhibit basilical layouts and date broadly to 730–867[24] This aligns perfectly with the period of King Offa, 757–796. Offa established an archbishopric at Lichfield in 787, elevating it above Canterbury. The apse is perfectly semicircular, extremely wide, without chapels or buttresses, basilical in form and is unmatched in English Norman cathedrals. A comparison shows no equivalent.   

   Comparison with English Norman cathedrals

  Rodwell suggested parallels with Jumièges or Bernay, but these reconstructions are conjectural and still do not match Lichfield’s proportions. A search of all Romanesque churches, 800–1200, as described by Krautheimer[25] and Conant[26] has failed to find a close affinity with the apse at Lichfield with the early build of Jumièges Abbey coming closest. There is a resemblance to St Denis Abbey, located in a suburb of northern Paris[27]. Abbot Fulrad built a basilica church, dedicated c.775, with many features modelled on St Peters in Rome. Partial excavation in 1938 by Crosby[28] revealed a wooden roofed columnar basilica with a spacious adjunct extending a little beyond the aisle walls, a lantern tower, a new kind of west end, and a simple, short, apse extending from the east end.   

 

Reconstructed Basilica of Saint-Denis, the earliest Carolingian Romanesque church. Nave and apse were 30 feet wide. The alignment of the crypt is unexplained and shape conjectured.

 

 A monumental basilica‑style cathedral would have been a fitting architectural expression of Mercian ecclesiastical ambition,[29] possibly inspired by Carolingian models such as Fulrad’s Saint‑Denis (dedicated 775).[30]

AI rendition of statue of King Offa on the west front of the cathedral.

 

Why the Norman Interpretation Fails

Victorian scholars, influenced by the grandeur of Norman cathedrals, assumed the foundation must be Norman. However, the metric is Early Medieval, the construction is far more likely to be Early Medieval, the proportions match Early Medieval basilicas, the absence of Norman masonry is conspicuous and the layout is unlike any Norman cathedral in Britain. Rodwell’s proposed Norman date of 1085[31] cannot be reconciled with the evidence.

          This conclusion would be placed beyond doubt with a Radiocarbon Dating. The foundation lies only 30 cm below the choir aisle floor. Two mortar samples could be extracted easily and dated for approximately ~£2.5k.

[1] John Hamlet’s plan showed the position of an old foundation is held in the Staffordshire Record Office LD 289/16. The drawing has a few revisions with pen overlaying pencil. Willis’s notes are kept in Cambridge, Cambridge University Library, MS 5042, with the title ‘Architectural drawings and notes of Lichfield Cathedral 1861 64’. Willis saw only the south east part of the apse. The greatest amount of the foundation wall was observed by Rawson and Hamlet in 1856 and 1860. Map reference for the apse is SK1159 0978.

[2]  R. Willis, ‘On foundations of early buildings recently discovered in Lichfield Cathedral’ The Archaeol. J. (1861) 18, 1–24. Willis visited the Cathedral in 1849 to examine window tracery. In 1854, he was invited to forward a drawing for the restoration of the choir area. Before publication Willis gave a lecture reported in The Gentleman’s Magazine, March 1861, vol 210, 296‑300.

[3] ‘Seemed to lie’ was Willis’s phrase. The gap between the lateral wall and cathedral wall was estimated to be c. 180 mm (7 in.). Hamlet’s drawing showed the bottom of the foundation was wider than the top.

[4] It has also been found in smaller churches, but often for one dimension, length or width, only.

[5] During the 1856 installation of heating ducts, men had to force a way through concrete of ‘unusual hardness’. J. G. Lonsdale, Recollections of work done in and upon Lichfield Cathedral from 1856—1894 (Lichfield 1895), 6.

[6] W. Rodwell, ‘An interim report on archaeological excavations in the south quire aisle of Lichfield Cathedral’, (Unpub. report held in Lichfield Cathedral Library 1992b),

[7] G. M. Livett, ‘Foundations of the Saxon Cathedral Church at Rochester’ Archaeologia Cantiana (1889) 18, 264.

[8] A.W. Clapham, English Romanesque architecture (Oxford 1930), 156.

[9] H. M. Taylor, Anglo-Saxon architecture: vol III (Cambridge 1978), 761 and 964.

[10] H. M. Taylor and J. Taylor, Anglo-Saxon architecture: vols I and II (Cambridge 1965), 7 and 1978, 756.

[11] J. Blair, ‘Grid planning in Anglo-Saxon settlements: the short perch and the four perch module’, in H. Hamerow ed., Anglo-Saxon studies in Archaeology and History, 18 (Oxford 2013), 54 and J. Blair, Building Anglo-Saxon England (Princeton and Oxford 2018), 71, 149.

[12] H. Braun, An introduction to English Medieval architecture 2nd ed. (London 1968), 71.

[13] J. Blair (2018) 160. See note 11...

[14] J. Blair (2013), 24. See note 11.

[15] Pre-Conquest walls are seldom as much as 3 ft in thickness and are more often nearer 2 ft 6 in.; whereas Norman walls are seldom less than 3 ft thick. However, there are exceptional Anglo-Saxon walls..

[16] This correlation applies to the apse foundation, but not the foundation under the choir which varies in width. The south foundation is wider than the north foundation and is probably because the ground slopes southwards.

[17] W. Rodwell, ‘Anglo-Saxon church building: aspects of design and construction’, in L. A. S. Butler and R. K. Morris eds., The Anglo-Saxon Church. Research Report 60, The Council for British Archaeology (1986), 157.

[18] See note 11, J. Blair, (2013), 26.

[19] The distance between the inner faces of the Romanesque column foundations under the floor of the nave were slightly narrower than the present columns.  

[20] There are two pier bases on the choir west wall drawn on Hamlet’s 1856 drawing

[21] H. M. Taylor, Anglo-Saxon architecture: vol III (Cambridge 1978) 1063. Taylor had little doubt that most Anglo-Saxon churches were plastered inside and outside.

[22]  D. Parsons and D. S. Sutherland, The Anglo-Saxon Church of All Saints, Brixworth; Northamptonshire: Survey, excavation and analysis, 1972—2010 (Oxford 2013).

[23] E. Fernie, The architecture of the Anglo-Saxons (New York 1983), 64–5.

[24] E. Gilbert, ‘Brixworth and the English Basilica’ in Art Bulletin, (1965) vol 47, 1, 14. Gilbert concluded English basilicas were erected between 730 and 867.

[25] R. Krautheimer, R. Early Christian and Byzantine architecture. (Harmondsworth: 1965).

[26] K. J. Conant, Carolingian and Romanesque architecture, 800–1200. (New Haven and London: 1978).

[27] The old church, c. 475, resembled Wilfrid’s church at Hexham according to Clapham (Oxford: 1930).

[28] S. M. Crosby, The Royal Abbey of Saint-Denis from its beginnings to the death of Suger, 475–1151. (New Haven and London: 1987).

[29] S. Shaw, The history and antiquities of Staffordshire: vol I (London 1798), 234; J. Jackson, History of the City and Cathedral of Lichfield (London 1805), 73–74 and T. Harwood, The history and antiquities of the church and city of Lichfield (London 1806), 7. 

[30] Gilbert 1965, 1, suggested Brixworth church was contemporary with Fulrad’s Saint-Denis church.

[31] Rodwell fixed a start of build in 1085 and described it as Norman Romanesque in style. Fragments of moulded masonry identified as Romanesque were found,  Rodwell 1992a, 29; R. K. Morris, ‘The lapidary collections of Lichfield Cathedral’, in J. Maddison ed., 13th Medieval Archaeology and Architecture at Lichfield (Leeds: The British Archaeological Association 1993), 101–108; Rodwell 1994, 29.