Fifteen of the stone monuments described in this volume carry inscriptions in Latin lettering. A further six have texts in runes and one of those with Latin lettering (at Orpington, Kent) also carries three probably runic characters of uncertain meaning (see the catalogue entry). The numbers of inscriptions in the south-eastern counties is low in comparison with the north of England (Okasha 1971, map I; Page 1973, fig. 7). This is probably due to the greater intensity of later development in the south.
Most of the surviving inscriptions in the south-east date from the last century or so of the Anglo-Saxon period. The early church has left little or no epigraphic trace. This is in marked contrast with Northumbria where important groups of inscriptions remain from several documented ecclesiastical centres, such as Lindisfarne, Jarrow, Monkwearmouth, Hartlepool, Whitby and York (Okasha 1971; Higgitt 1979). This strongly monastic phase of the early church is unrepresented among the surviving inscriptions of the south-east. Even Canterbury, whose early inscriptions might tell us much about the cultural contacts of the early church there, remains a blank apart from a fragmentary inscription of uncertain date from St Martin's and a very late grave-cover from St Augustine's (no. 2). The recorded texts of verse epitaphs of three early archbishops, however, suggest that inscriptions were taken very seriously in early Canterbury (see below).
Although there are three inscriptions in Latin lettering from pre-Conquest Winchester and two each from London and Canterbury, each of the inscriptions is sui generis and no epigraphic centre with definable characteristics comparable with the Northumbrian centres mentioned above emerges in the south-east.
Where their origin is clear the inscribed stones are architectural features, stone crosses, or grave-markers. There are four architectural inscriptions. Two of these are sundials (Bishopstone and Orpington (Ills. 6–7, 105–7)), which may be compared with the group of late Anglo-Saxon inscribed sundials in Yorkshire (Higgitt in Lang 1991, 123–4, 133–5, 163–6, 195, ills. 418, 451–3, 568–73, 729–31). A fragment of a dedication inscription, which is probably re-set in its present position on a door jamb of the early chancel at St Martin's in Canterbury (Ill. 57), can be compared with a number of other Anglo-Saxon inscriptions of various forms that record details of the dedication of churches (Higgitt 1979, 346-7, 368-9). The prominent inscription on the extrados of the arch that opens from the crossing to the transept of the later tenth- to early eleventh-century church at Breamore is unusual both in the size of its lettering and in its bold and disciplined architectural setting (no. 2c; Ills. 429–30, 433–7). The letters, which are in the region of 15 cm high, and the three letters from another inscription re-set over the chancel arch, which seem to be even larger (Ill. 431), are very probably the largest to survive from Anglo-Saxon England. The setting of carved lettering on the extrados of an arch is without parallel in Anglo-Saxon England and seems to be very unusual elsewhere in the early Middle Ages. The idea was perhaps suggested by lettering in mosaics in Rome, where inscriptions can still be seen on the extrados of the triumphal arches in San Paolo fuori le mura and San Lorenzo fuori le mura and of the arch around the apse in the Triclinium of Leo III (Oakeshott 1967, pls. 77, 112, 183, 185). Certainly such monumental and architectural treatment of lettering is very unusual in the earlier Middle Ages (Petrucci 1980, 5–9; Mitchell 1990, 205–16).
The two inscriptions from All Hallows by the Tower in London are on fragments of stone crosses (Ills. 320, 324, 327, 343–4). One (no. 2) is inscribed around a head (Ills. 343–4) and the other (no. 1) within the sculptural field on the shaft (Ills. 320, 324, 327). Inscribed stone crosses of the Anglo-Saxon period are quite common in the north of England but only three remain in the south (Higgitt 1986, 147–8).
Five of the other inscriptions (from St Augustine's Canterbury (no. 2), Rochester 3, Stratfield Mortimer, Whitchurch, and the Old Minster cemetery in Winchester) are on grave-covers and head- or foot-stones (Ills. 24–8, 147–50, 695–709, 482, 485–9, 509–13, 521). The two other Winchester fragments with Latin lettering may also have come from funerary monuments but the original form of these monuments is uncertain (Ills. 490–3, 676–8). The monuments with runic inscriptions from Dover (St Peter), London (St Paul's cathedral) and Winchester (St Maurice 1) were also grave-markers or -covers and those from Rochester (no. 2) and Sandwich (no. 1) in Kent are also likely to have been so.
The texts seem to have served a variety of different purposes and to have spoken on behalf of different sorts of patrons. Most had as one of their functions the commemoration of individuals (the dead, patrons and makers). The language may tell us something about the intended audience, or at least about the patrons. Probably seven of the extant monuments have texts in Latin (Canterbury (St Augustine's 2 and St Martin's 1), Orpington, Stratfield Mortimer, Whitchurch, and Winchester (Old Minster 1 and Lower Brook Street 1), and at least four, probably five, have Old English texts (Breamore 2, London All Hallows 1–2, Orpington, and Winchester Old Minster 6). Two of the runic texts are also demonstrably in the vernacular, in their cases Old Norse (London St Paul's cathedral and Winchester St Maurice 1).
The four architectural inscriptions have little in common with each other. The Bishopstone sundial displays a name, presumably that of a patron or benefactor, without further comment. The Old English and Latin texts on the Orpington sundial probably refer to the function of the sundial, as do texts on sundials at Kirkdale and Great Edstone in Yorkshire (Higgitt in Lang 1991, 133–5, 164–6). The Orpington texts seem to betray some pride in the possession of a sundial and in the computistical knowledge needed to read the dial; and the obscure runes also hint at pleasure in special knowledge. The inscription on St Martin's in Canterbury (Ill. 57) seems to have been in Latin and to have recorded (in accordance with church law) details of the dedication, apparently with the common formula in honore(m) NN (Higgitt 1979, 346–7, 368–70). The Old English text of the Breamore inscription is uniquely prominent in its display but unfortunately its purpose is not clear. The language is reminiscent of some Old English charters but as the text could be incomplete at either end it is difficult to know whether the covenant or agreement it refers to is a covenant with God or a record of an earthly transaction, although a good case for the former has recently been argued (Gameson and Gameson 1993). This (now) uniquely prominent display of a text within an Anglo-Saxon church perhaps makes a religious meaning more likely (Ill. 429). To record a major benefaction or privilege in this way would be without parallel among Anglo-Saxon inscriptions but charter-like notes of donations of land are displayed in inscriptions on stone elsewhere, for example on crosses of the tenth or eleventh centuries at Merthyr Mawr and Ogmore in Glamorgan and in a probably ninth-century inscription at Santa Maria Maggiore in Rome, which was carved and displayed 'pro cautela et firmitate temporum futurorum' (R.C.A.H.M.W. 1976, 55–6, 57; Gray 1948, 100-1).
The head of the one of the two crosses from London (All Hallows 2) carries a memorial inscription in Old English. Memorial texts are the most frequent type of text to be found on Anglo-Saxon crosses (Higgitt 1986, 133–4). As often in the earlier Middle Ages, the inscription apparently recorded both the name of the deceased and that of the person who had the monument raised (Gauthier 1975, 43–8; Higgitt 1986, 133, 139; Moltke 1985, 189–91, 224–41, 289–327 passim). The approximately contemporary runic inscription on the grave-marker from St Paul's in London (Ill. 350) similarly records, in Old Norse, the names of those who raised the monument (see the catalogue entry). The other London cross (All Hallows 1) has an incomplete text of uncertain meaning but probably also in Old English, cut in the spaces by the legs of a carved figure. The position of the text suggests that it may have referred to the sculpture. When Leland saw the Reculver cross in the sixteenth century, it too was inscribed (Leland 1964,
Three of the grave-markers are inscribed in Latin and three further fragmentary inscriptions, two in Latin and one probably so, probably also come from grave-markers. The complete texts at Canterbury (St Augustine's 2), Stratfield Mortimer, and Whitchurch, use known Early Christian and medieval memorial formulae or variants of them. All three identify the burial. The two late, probably eleventh-century, grave-covers (Canterbury St Augustine's 2 and Stratfield Mortimer) further identify the deceased by naming his father. The former notes the date (but not the year) of death, which would have aided annual commemoration in the same way as entries of obits into an obituary or calendar. The date noted at Stratfield Mortimer, on the other hand (that of the burial and not the death), is of less obvious use (cf. Gauthier 1975, 49). These two inscriptions also include requests for prayers for the soul of the deceased but use quite different formulae to do so. The Canterbury request contains an early use of the requiescat in pace formula, which, Rieckenberg has argued, originated in the liturgy of Mainz in the tenth century (Rieckenberg 1966). This may have encouraged the usage but there seem to have been some earlier epigraphic examples (Jörg 1984, 113). Finally these two late grave-covers have a further point in common. Both conclude with a maker formula of the common 'speaking object' type: me fecit in Canterbury and me scripsit at Stratfield Mortimer. The verb scripsit implies literacy and probably refers to the drawing up of the text or to the design of the lettering (see below). The commemoration of the craftsman and/or designer in this way can be compared with similar formulae on more or less contemporary sundials in Yorkshire (in Old English at Great Edstone and Kirkdale and probably in Latin at Old Byland) (Higgitt in Lang 1991, 46, 134–5, 164–6, 195).
The two fragments from Winchester (Old Minster 1 and Lower Brook Street 1) have what are probably parts of Latin memorial formulae (hic and vivat in evum). All that is legible on the Rochester 3 inscription is amen, in all probability the conclusion of a memorial prayer. The language of the prayer is, however, uncertain.
The grave-cover from the cemetery of the Old Minster in Winchester (no. 6) is in Old English. Its her līš formula corresponds to the common Early Christian and early medieval hic requiescit which was used in England at Monkwearmouth, co. Durham, and on Whitchurch 1, and was recorded by Bede in epitaphs at Canterbury and Ripon (Bede 1969, 144–5 (
There are some non-palaeographical indications of date of various sorts for some of the inscriptions. The St Martin's Canterbury inscription is built into the fabric of the early (perhaps seventh-century) chancel, but it seems not to be an original feature (see Chap. V). One of the fragments excavated in Winchester (Old Minster 1) can be dated on archaeological grounds to before c. 901–3. The head-stone at Whitchurch has been attributed on stylistic grounds to the ninth century (see Chap. V and the catalogue entry). The archaeological contexts of two other inscriptions from Winchester also provide dating evidence for them: the early eleventh century for the tombstone of Gunni and some time before about the middle of the eleventh century for the HIC fragment. The church at Breamore, which is agreed to belong to some time between the middle of the tenth and the early eleventh century, provides a terminus post quem for its inscription.
Scandinavian influences provide another class of dating evidence. The ornament in the Ringerike style on Rochester 3A (Ill. 147) is very close to that on the rune-inscribed grave-marker from St Paul's in London (Ill. 351); an early eleventh-century date is likely for both (see the catalogue entries). The tombstone of Gunni in Winchester (Old Minster 6) has been dated to late in the pre-Conquest period, and probably to the Danish period, because of the Scandinavian name and the use of feolaga, which is in origin an Old Norse loan-word, and perhaps also that of eorl as a title (Kjølbye-Biddle and Page 1975, 391–2; Page 1971, 180; see the catalogue entry). The Old English memorial formula on the cross-head from London (All Hallows 2) appears to have been influenced by wording sometimes found in Norse inscriptions (Okasha 1967; Page 1971, 178). One of the reasons for seeing the tombstone at Stratfield Mortimer as late is the Old Norse origin of the name of the craftsman, author, or designer of the text.
What determined the choice between Latin lettering and runes for inscriptions in this area? Although the runic inscription from Sandwich (no. 1; Ill. 156) has been dated to as early as the seventh century, its meaning and date are uncertain (see the catalogue entry). The earliest reasonably certain dating for any of the inscriptions in Latin lettering is to some time before c. 901–3 for one of the inscribed fragments from Winchester (Old Minster 1). There is no useful evidence for answering the question for the earlier part of the period. The introductory cross and name inscribed in Anglo-Saxon runes on the tombstone from Dover St Peter (Ill. 76) are well cut and lightly seriffed in a manner that implies that the letter-cutter had experience of cutting inscriptions in Latin lettering (cf. Page 1973, 104, fig. 25). There may have been some knowledge of both Anglo-Saxon and Scandinavian runes at Orpington, although the significance of the three runes, if that is what they are, is obscure, and they may be secondary to the texts in Latin lettering. Two inscriptions in Old Norse and inscribed in Danish runes on what were probably funerary monuments (London St Paul's and Winchester St Maurice 1) have been attributed to the early eleventh-century period of Danish control (Kjølbye-Biddle and Page 1975; Page 1971, 175). The patrons seem to have been members of the Danish elite. Gunni probably belonged to the same social group, but was commemorated instead in Scandinavian-influenced Old English in Latin lettering, a sign, presumably, of greater integration with the indigenous culture. The Ringerike stone from Rochester (no. 3), which is also inscribed in Latin lettering, may show the same thing, if it is not a case of the Anglo-Saxon adoption of Scandinavian styles.
The lettering of the non-runic inscriptions is, with one exception, in capitals which take the forms of Roman capitals or of common variants of them (see Note on Capitals below). The exception is the fragmentary inscription on the cross-shaft from London (All Hallows 1) (Ills. 323–4, 327). Only two of the five different letters of the Latin alphabet approximate to their Roman capital form. The other three forms are uncial D, probably half uncial H, and oblong O. This eclectic mixture of forms is reminiscent of the 'decorative capitals' found in the display script of many Insular manuscripts of the seventh to ninth centuries and in a number of contemporary inscriptions on stone, especially in Northumbria (Higgitt 1982, 310–15; idem 1994). This resemblance is reinforced by the use of wedge-like serifs, which were characteristic of early Insular lettering, on the London cross-shaft. This inscription is the only evidence that lettering reminiscent of Insular decorative capitals was used on stone in the south-east of England. The late character of the decoration on this cross means that this inscription must be a late reflection of Insular decorative capitals. Such lettering would, however, be surprising much after the end of the ninth century. Certainly the decorative capitals that appear in the display script of one or two English manuscripts of around 900 are only dimly reminiscent of earlier Insular capitals (Temple 1976, ills. 3 and 9) and Roman capitals are the normal display script in English manuscripts of the tenth and eleventh centuries.
As was seen above, very little is known from surviving material about the inscriptions of the first two to three centuries of the Anglo-Saxon period after the conversion to Christianity. Three of the inscriptions might be datable to before c. 900: one of the Winchester fragments (on archaeological grounds to before c. 901–3); the inscription at St Martin's in Canterbury (uncertainly associated with the early Anglo-Saxon fabric of the church); Whitchurch (attributed on art-historical grounds to the ninth century). All three use Roman capitals and variants of them (Ills. 57, 485–9, 490). None of the forms are sufficiently characteristic to help in dating. The wedge-like seriffing on the Whitchurch inscription is an embellishment that would be unlikely to be used much after the ninth century (Ills. 485–9). These three inscriptions suggest that Roman capitals had been in use for inscriptions in the south-east for some time before c. 900. It is quite possible that they were the normal epigraphic script from the time of St Augustine. They had been one of the epigraphic scripts in Northumbria from the later seventh century on (Higgitt 1979). It is, however, not unlikely that there was also some use of Insular decorative capitals. These might have resembled those in the display scripts of manuscripts of the eighth and ninth centuries which were probably written at Canterbury, where the 'decorative' forms appear alongside a high proportion of Roman capital forms (Alexander 1978, ills. 144, 145, 152, 160).
The other source for the early Anglo-Saxon epigraphy of the south-east is in the recorded texts of the long Latin verse inscriptions that served as epitaphs for archbishops of Canterbury in the late seventh and earlier eighth centuries. They were composed in imitation of the ambitious verse inscriptions of the continental and especially the Roman church. The texts of such inscriptions were known through manuscript collections and some of the epigraphic verses from England were in their turn copied into books (Lapidge 1975). Bede quoted the first four and last four verses of the thirty-four verse epitaph marking the burial of Archbishop Theodore ('Hic sacer in tumba pausat cum corpore praesul ...') in the church of Saints Peter and Paul in Canterbury (Bede 1969, 474–5 (
In the sixteenth century Leland transcribed a number of such verse texts from an eighth-century manuscript collection from England and these included the epitaphs of two later archbishops of Canterbury, Berhtwald, who died in 731, and Tatwine, who died in 734 (texts in Lapidge 1975, 810–12). The epitaphs of all three archbishops were in elegiac couplets, as also were the tomb-inscriptions in Rome of Pope Gregory I (d. 604) and Caedwalla (d. 689), which were known to, and recorded by, Bede (Bede 1969, 132–3 (
The same collection includes eight lines of rhythmic and rhymed octosyllables commemorating a dedication of a church or an altar to St Paul by Bishop Haedde of Winchester (676–705) (Lapidge 1975, 817). The verses open with the in honorem formula. It is not clear whether these lines were taken from an inscription or whether they were simply a manuscript exercise in the genre but in either case these lines, along with the Canterbury epitaphs, show that the early church in the south-east was familiar with the tradition of Latin verse inscriptions.
The other inscriptions included in this volume all probably belong to the period after c. 900 and all use Roman capitals, although there are some variations in style. The inscriptions at Breamore (no. 2) and Orpington are well-preserved examples of neat and well executed lettering and seem to be characteristic of one late Anglo-Saxon style (Ills. 105–7, 429–33). The letter strokes are of even breadth with no contrast of thick and thin and they have no noticeable seriffing. The letters tend to be tall in their proportions. The forms of the capitals are Roman but one or two angular variants occur (C, G and S) and, as often, slight differences are introduced in the treatment of the cross-bars and tops of A. The Breamore inscription is probably contemporary with the arch, which was probably built within a decade or two of the year 1000. The Orpington lettering is likely to be of approximately the same sort of age. The inscription on Gunni's tombstone in Winchester (Old Minster 6 (Ills. 511–13)) is generally similar in execution and forms (angular G). The two or three letters left on Rochester 3 may also have belonged to an inscription in the same style (Ills. 149–50).
The rather worn lettering on the inscription on the tombstone at Stratfield Mortimer is a little squatter and is not so even (Ills. 698–709). Again the strokes are of even breadth but they finish in serifs. The capital forms are Roman with the exception of angular C and G, uncial Q and variations on A. The lettering cannot be closely dated but it would fit well with the eleventh-century date suggested by the contents of the inscription.
The neat capitals of the Breamore/Orpington style were probably contemporary with the capitals used in the display script of many southern English manuscripts in the period following the tenth-century Benedictine Reform (see Temple 1976, passim, for numerous illustrations; Heslop 1990, 163–6). They have the same regularity and their angular variants can be found in some of the manuscripts (Temple 1976, ills. 114, 115, 126, 133, etc.). The even and unseriffed lines of the inscriptions seem, however, to be an epigraphic development and are quite unlike the contrasting thicks and thins and the serifs of the manuscript capitals.
The other four inscriptions are distinctive in various ways. The inscription on the cross-head from London All Hallows (no. 2) is more roughly designed than the others (Ills. 343–4). Again the underlying forms are Roman capitals but the rune-like angular R and the extension of the verticals of E and L below the bottom horizontal and perhaps (in the case of E) above the top horizontal look like barbarisms in comparison with the regular, almost Classical forms of Breamore and Orpington (Ills. 105–7, 429–33). The probable Scandinavian influence on the Old English text accords with such lettering and it also suggests a date around the last century before the Conquest. This lettering is also exceptional in the use of dot serifs. These were also used to embellish at least some of the letters of the Bishopstone inscription (Ills. 6–7). Here the capitals are more regular in their forms and execution but are not closely datable.
The fragment from Winchester with the letters HIC (Lower Brook Street 1) does not offer much scope for palaeographical dating (Ill. 678). It is of interest because of its broad letter strokes which flare out towards the ends and because of the cross-section of the trenches which would only make sense if they were intended for some kind of inlay. The idea may have been suggested by lead-filled lettering such as that of the Carolingian epitaph of Adalberga in Tours (Gray 1986, ill. 69).
The lettering of the grave-cover from Canterbury (St Augustine's 2) is made up of slender, even strokes which terminate in small serifs (Ills. 25–8). The capitals are Roman. No angular forms are used; on the other hand a rounded (uncial) E is used alongside the normal capital form. Tall I is a little surprising; it is a feature that could have been imitated from Roman inscriptions. The lettering is likely to be later than the Breamore/Orpington phase, and similarities to that on the Bayeux Tapestry suggest that it could date from shortly after the Conquest.
This region provides an impressive amount of evidence for the use of colour in conjunction with pre-Conquest inscriptions on stone (both in Latin lettering and in runes). There are records or surviving traces of colour on more than a third of the inscribed stones discussed in this volume: at Breamore 2, London All Hallows 2, London St Paul's 1, Reculver 1, Rochester 2 and 3, Stratfield Mortimer, and Winchester St Maurice 1 (Tweddle 1990, 150–1). In the case of Rochester 3, the visible colour is not in or around the letters, but elsewhere on the stone (Pl. 1; Ills. 147–50). As Tweddle points out, the colour may in some cases be secondary (Tweddle 1990, 150). The amount of evidence for colour found in this sample of inscriptions confirms the view (cf. Higgitt 1986, 131–2, 138–9, 143) that it was usual to pick out carved letters in colour.
There are cracks in the normally anonymous facade of inscriptions in one or two of this group of inscriptions. The maker formulae at Stratfield Mortimer (me scripsit) and St Augustine's in Canterbury (me fecit) could refer to all, some or just one of the various aspects of the production of the inscriptions: the drafting of the text; the designing of the inscription, not necessarily on the stone itself; the laying-out of text on the stone; the cutting of the lettering; and the painting of the inscription. Some craftsmen no doubt needed the assistance of a literate person with the wording and the letter forms (Higgitt 1990, 151–2). Others were perhaps literate enough to design and carve a text. The development of an epigraphic style of lettering like that at Breamore and Orpington shows that the cutting of inscriptions could be a routine and professional matter. The correction of a wrong letter in the Whitchurch inscription (Ill. 487) and the probable additions in the St Augustine's Canterbury (Ill. 25) inscription would be explained either by a literate letter-cutter thinking about the text while working on it or else by close supervision from a literate person. The Canterbury inscription stitches together epigraphic phrases, in one case apparently without fully understanding the syntax of the model. Was this composed by a professional stone-cutter with a stock of formulae or by an amateur monastic collector of epitaphs?
NOTE ON CAPITALS
The capitals used in Anglo-Saxon inscriptions vary considerably and it is not possible to isolate a single distinctive Anglo-Saxon type of capital (Okasha 1968; Okasha 1971). For the purpose of description and analysis it is helpful to compare examples with the forms of 'Roman capitals'. Roman capitals were made known throughout the Roman Empire through their use in inscriptions. Early Christian and early medieval book and epigraphic scripts developed away from this capital script but Roman capitals continued to exert an influence, especially in Italy. At certain periods and in certain places there seem to have been attempts to reproduce comparatively 'pure' Roman capitals in inscriptions and manuscript display scripts, most famously during the Carolingian period (Gray 1986; Bischoff 1990, 55–61).
When reference is made in discussion of letter forms to 'Roman capital' forms, this is not meant to imply a conscious imitation of Roman forms, although it is possible that 'purer' or 'more Classical' Roman forms were sometimes deliberately adopted because they were thought to look more 'Roman'. I have argued that this was the case at Monkwearmouth and Jarrow, co. Durham, in their early years (Higgitt 1979).
To avoid ambiguity the 'Roman' or 'Latin' alphabet is here referred to as the 'Latin alphabet', in order to distinguish it, regardless of letter forms, from other alphabets such as runes.18
NOTES
17. I would like to thank my daughter, Catherine Higgitt, for help during the field-work undertaken in preparing the non-runic epigraphic descriptions in this volume. I am also very grateful to David Parsons for his helpful comments on my discussions of Old English texts and names and also for his contribution to the discussion of the probable runes at Orpington.
18. Following standard international practice, the typographical convention of transcribing Scandinavian runes in lower case bold without inverted commas has been adopted throughout this volume (Eds).
