Lost

Found in 1770 at Castle Hill, Caistor by labourers digging materials for the repair of the roads (Turner and Bradley 1784, 161–2; Trollope 1861–2, 152). The stone had been removed to 'the Museum at Lincoln' by 1831 (Oliver 1831, 204); at that date this is likely to mean the collection in the Cathedral cloisters, some of the portable objects from which became the basis of Lincoln Museum in 1906, rather than the miscellaneous items held by the Mechanics Institute after its foundation in 1833 in the Lincoln Greyfriars. However, an undated record in H. E. Smith's antiquarian notebooks on Caistor (Notebook 108B, 46) comments, 'Unfortunately this stone is not preserved, but a rough cast of it is in the possession of Mr John Witham'.

Castle Hill

Unknown

See Barton-Upon-Humber 4.

Inscription If Radford (1946, 95–6) was right to accept Turner's statement that the letters were approximately 4 inches high (10 cm), this would have been an unusually monumental inscription by surviving Anglo-Saxon standards. The 5–5.5 cm of the lettering of the Jarrow dedication inscription would be much more normal. It certainly seems to have been an impressive inscription with its moulded border (see below) and its ornamental lettering. As depicted in the engravings published in 1784 (Ill. 79) and 1789 (Ill. 80) the strokes of the letters are broad with little variation in thickness. [Note: The 1789 engraving is clearly based on that of 1784 and so is not an independent witness. Where the two differ, for example where the plate of 1789 comits the cross-bar of the first E in line 3 as shown in the earlier engraving, it is presumably less accurate.] They appear to have flared out gently towards the ends into trumpet-like terminations. Dr Radford has made the interesting suggestion (pers. comm.) that the Caistor inscription was in raised lettering or false relief. This was a laborious technique which was rare in Insular lettering but not unknown, as a handful of examples of the eighth and ninth centuries at Tarbat, Wensley and Bealin show (Higgitt 1982, 315–17). Against this it could be objected that finer strokes like the leg of the first R, if correctly shown, would be difficult to execute in relief and also that vulnerable relief lettering would probably have suffered greater damage given the condition of the stone. If the strokes were indeed incised and as broad as they appear to have been, the best Anglo-Saxon parallels might be the fragment of incised lettering from Winchester and the lost fragment from Shaftesbury (Okasha 1983, nos. 175, 183; Tweddle et al. 1995, 331–3). It looks as if the Winchester letters were designed for some kind of inlay. It is just possible that this was the case at Caistor.

The forms of the lettering seem to have been copied with some care in the eighteenth-century engraving (Ill. 79). The lettering is strikingly similar in proportions, flared trumpet-like terminations, and individual letter forms to some of the more angular examples of manuscript display script using Insular decorative capitals. The Collectio canonum in Cologne and the Lichfield Gospels have passages of display script that are very reminiscent of Caistor in their treatment and their rectangularity (Radford 1946, 96–7; Alexander 1978, ills. 60, 76 and 78). Insular decorative capitals are a script composed largely of variations, which are often angular, on Insular half uncial and capital forms. They first appear as a mature display script in the Lindisfarne Gospels (Kendrick et al. 1960, 75–7, 93). They flourished as a manuscript display script during the eighth century and they also appear in inscriptions on stone in a number of Insular centres, particularly in Northumbria, at about the same time (Higgitt 1994).

An analysis of the decorative capitals of the Caistor inscription bears out Radford's conclusion that it 'should belong to the eighth century' or perhaps to the first half of the following century (Radford 1946, 97). Most of the individual letter forms can be matched in Insular manuscripts and inscriptions of this time; the others are elaborations of known forms. For parallels to several of the letters reference will be made to my analysis of the display script of the Book of Kells, without meaning to imply any particular kinship with the Book of Kells itself (Higgitt 1994, especially 219–33). The Es, Is, Ns, Rs, and the rectangular U in line 3 correspond to Kells E2, I, N1, R1 and U/V2. The rectangular Cs are elaborations of Kells C; and the H is a version of Kells H1 in which the second stroke is inclined rather than horizontal. B seems to have been a hybrid form of the capital, rounded above but with a rectangular base. This was either intended by the designer to assimilate the letter to its neighbours or it was an error by the eighteenth-century copyist. Capital B appears, for example, in Durham Cathedral MS A.II.17 (Alexander 1978, ill. 47) and on the Ruthwell Cross (Cassidy 1992, pl. 16). The D is a rectangular variant of the half uncial letter which can be matched in St Gall Codex 1395 (Alexander 1978, ill. 261) and in Northumbrian inscriptions at Auckland St Andrew, Hartlepool and Wensley (Okasha 1971, nos. 11, 45, 46, 120). The sixth letter in line 3 was probably a rectangular form of capital or uncial G. Q is comparable to the rectangular letter in the Lichfield Gospels (Alexander 1978, ill. 78). The Roman capital V in line 2 is not very common amongst Insular decorative capitals, although there are examples in books and on stone (Higgitt 1994, 232–3). The very unusual rectangular form of the half uncial S in line 2 also appears in the early ninth-century Book of Armagh, which also has a simpler form of the rectangular P that can be seen at Caistor (Higgitt 1994, pl. 74). The most striking letter is the rectangular O, the verticals of which are extended above and below the horizontals. This can be found occasionally in manuscript display script: in the Cologne Collectio canonum; in London, British Library, Royal MS 1.B.VII; and in the Lothian Psalter (Alexander 1978, ills. 60, 73, 149). Interestingly the Lothian Psalter was in Lincoln at the end of the Middle Ages but its display script is softer, more rounded and more reminiscent of southern English manuscripts than the lettering of the Caistor inscription. Caistor seems rather to belong with the northern tradition represented by books such as the Lindisfarne Gospels, the Cologne Collectio canonum and the Lichfield Gospels. The sophisticated lettering of the Caistor inscription was in all probability designed by someone familiar with manuscript display script and it suggests close links with a centre of manuscript production, perhaps at Caistor itself. The erased lower script on the lead tablet recently found at Flixborough seems now to be largely illegible but it shows that angular decorative capitals were known elsewhere in eighth- or early ninth-century Lincolnshire (Brown 1991; Okasha 1992a, no. 193). The taste for Northumbrian angularity was probably fading south of the Humber before the end of the eighth century (Brown 1991). It is reasonable to conclude that the decorative capitals of the Caistor inscription belong most naturally in the eighth century.

The apparent lack of incised horizontal lines between the lines of text is another feature that suggests that this was a comparatively early inscription (Higgitt 1995, 233; cf. also this volume, p. 215).

Function Assessment of the inscription and its detailed letter forms is subject to reliance on the accuracy of the eighteenth-century engraving. Its content, too, was associated from its very first recording with one specific historical interpretation, namely the supposed victory of Egbert of Wessex over Wiglaf of Mercia in the close vicinity of Caistor in 827 (Turner and Bradley 1784). So pervasive was this belief from an early date after the discovery of the inscription that Smith, for example, records that the same John Witham who had a cast of it also had a signet ring 'bearing the initial W, richly engraved, supposed to have belonged to Withlafe' (Notebooks 108B, 46). In fact, the Anglo-Saxon Chronicle's account of this conquest of Mercia does not mention or locate a decisive battle (Earle and Plummer 1892, s.a. 829). The roots of the association with Caistor lie solely in antiquarian speculation, and the possibility exists that the received reading of the inscription is itself biased by that presumed association. In practice, it cannot be sustained either by the documentary evidence or, as Radford (1946) has made plain, by what can be made of the inscription.

While acknowledging the need to treat the transcrip-tion cautiously, nevertheless the recorded combination of letter forms place the inscription comfortably in an eighth- to ninth-century context. They set it apart, for example, from the purely Roman capitals of the Jarrow dedication stone of the late seventh century and make it more akin to other Northumbrian inscriptions, that admit more Insular letter forms whose analogies exist in the display scripts of manuscripts (Higgitt 1979, 354–64; Higgitt 1994). In addition, the form of the substantially complete name E[G]BEREC– compares closely with that of other northern/non-West Saxon renderings of dithematic masculine personal names incorporating the common element -beorht in eighth- to ninth-century inscriptions, for example Falstone, Jarrow II, Lindisfarne I, Monkwearmouth II, Wensley II, Wycliffe, Yarm (Okasha 1971, nos. 39, 62, 75, 92, 121, 144, 145). The form EONBERECH[T] in the recently discovered local example of Flixborough II reinforces this (Okasha 1992a, nos. 193 and 208). According to Okasha's assessment of the latter piece (1992a, 46–7), the intrusive final E in –BEREC– may suggest a date 'not later than the ninth century'.

Radford also suggested that this stone formed part of a titulus or dedication stone of a church or altar within a church, and this seems probable. The clearest indication is the use of the dedicatory formula in honore(m), found also on the wooden altar in the relics of St Cuthbert of c. 698 and a fragmentary stone inscription from St Martin's church at Canterbury (Higgitt 1979, 347, 368). Both these Latin texts belong early in the Christian Anglo-Saxon period. A comparable formula occurs locally at St Mary-le-Wigford 6 in Lincoln in the later eleventh century, but in the OE form to lofe, and perhaps a little later again on the sundial from Stow (no. 6), also in OE. In addition, the Caistor inscription is set in a panel within an unusually elaborate border. Longer non-runic display inscriptions are typically organised in a panel bounded with some form of incised or moulded border, whether simply on a shaft as at Hackness, Hexham III, Wycliffe and Yarm (Okasha 1971, nos. 42, 54, 144, 145) or as part of sundials at Kirkdale and Skelton (ibid., nos. 64, 110). Further examples with raised or moulded borders are at St Mary Castlegate in York and Beckermet (ibid., nos. 8, 146). In a few cases, the thinness of the stone may indicate that the piece was a wall-fixed plaque (?Falstone, Thornhill, Whitby DCCXXXII and DCCXXXIII; ibid., nos. 39, 116, 133, 134). But particularly if the parallel lines of the engraving represent raised mouldings rather than incised lines on the original the effect at Caistor is akin to panelling within Roman epigraphic traditions. Whether a case of reuse, as at Lincoln St Mary-le-Wigford 6 (Ills. 273–5) and York Minster 42 (Lang 1991, 75–6), or emulation, it underlines the pretensions of the piece.

Those who saw it judged that the stone had been part of a wall-fixed panel, apparently because of its thinness. It presumably resembled, in function if not precisely in form, another dedicatory inscription that appears to have existed in Lindsey at a comparable date, which named Cyneberht bishop of Lindsey who died in 732 (Lapidge 1993, 2–3). This is discussed below under Unknown Provenance 1.

Context The occurrence of this dedication at Caistor confirms the presence of an early, presumably stone, church there in the pre-Viking period, but does not thereby automatically guarantee its pre-eminent status in Lindsey or its identification as the see of Lindsey as some have inclined to urge. A plausible context, as suggested by the Ordnance Survey (OS record card TA 10 SW 27, citing as authority C. W. Phillips's Dark Age index, cf. Ordnance Survey 1966), would be a pre-Viking monastery set within the late Roman stonewalled enclosure at Caistor (Rahtz 1960). The extant church of St Peter and St Paul stands within that enclosure and there is some evidence for the former existence of another church, dedicated to St Mary and situated a short distance (50m or less) to its north or north-east and also within the enclosure. Smith (Caistor notebook 'Antiquarian gleanings', p. 6) reports that: 'It was a popular tradition at the beginning of the nineteenth century that there was once a church of St Marie standing to the east of the Grammar School on which was then an orchard lawn, but from which every trace of ancient building had vanished'. Later thirteenth- and early fourteenth-century personal names also refer to 'Sayntmariland' (English Place-Name Society records). This suggests a family of churches within an enclosure typical of early monasticism. By Smith's account, too, every ground disturbance in the mid and later nineteenth century, not least the laying of the town's sewers, demonstrated that 'the whole ground once contained within the fortifications at Caistor is a veritable golgotha' (Notebooks, 1878). Among the (predominantly Roman) finds provenanced within the enclosure, a silver sceatta of Eadberht (737–758) of Northumbria and three further sceattas are direct evidence of eighth- and early ninth-century occupation (Blackburn 1993, 87). There is also an antiquarian record of 'a coin of the Byzantine series, struck in Nicomedia, too imperfect to be identified', found in 1858 (H. E. Smith Notebooks 108C, 35). Less certain is the identification of the location of a council of the bishops south of the Humber in 839, called 'Aet Astran' in the contemporary document, with Caistor in Lindsey (Birch 1885, 590–1 and note; Sawyer 1968, 403–4, no. 1438). In making this identification, Birch evidently picked up, via Lewis (e.g. 1845, I, 455), the firmly established interpretation that the Caistor inscription commemorated a victory of Egbert of Wessex over Wiglaf king of Mercia in 827 in the vicinity of Caistor.

Eighth or early ninth century

Turner and Bradley 1784, 161–2, fig. after 284; Gough 1789, II, 276 and fig.; Gough 1806, II, 386–7 and fig.; Oliver 1829b, 221–2; Oliver 1831, 204; Allen, T. 1834, II, 225–6; Lewis 1845, I, 445; Trollope 1861–2, 152; Binnall 1934, 7; Radford 1946, 95–9, fig. 3; Binnall 1960, 6–7; Okasha 1971, 58, no. 18 and pl.; Higgitt 1979, 347, 368; White 1980, 8–9 and fig.; Pevsner et al. 1989, 207; Stocker 1993, 117