The page as monument: epigraphical transposition in the runica manuscripta tradition of early Medieval England

: Most surviving runic inscriptions from early medieval England were produced in an ecclesiastical context, and the influence of manuscript writing practices on the runic tradition can clearly be discerned. The manuscript record of runes or runicamanuscripta thatflourishedparticularlyinthecontextofAnglo-Saxonmission-ary activity to the Continent has, however, usually been regarded as a late antiquarian development, largely detached from the epigraphical tradition. In this paper, I arguethatnotonlydidmanuscriptpracticeclearlyinfluenceepigraphy, butalsothat several uses of runes in manuscripts can be considered as extensions of the epigraphical tradition. Some runicamanuscripta also seem to evoke pointedly the monumental tradition, including associations with permanence, public display, and memorialisation. Through the case studies of decorative uses of runes, scribal signatures, and textual interventions in runes, I argue that there is a relatively consistent association between the runic script and monumental epigraphy that can be transposed onto the manuscript page for particular effects, which rely on received knowledge of the epigraphical tradition long after the use of runes in monumental contexts had ended.


Introduction
This paper takes a slightly different approach to the question of transposition between the epigraphical and manuscript traditions, focusing as it does on the ways in which the properties of monumental epigraphy-including permanence, the use of durable materials, and the association with ornamentation and memorialisation-found their way onto the manuscript page. By transposing features of epigraphy into a manuscript, the scribe was able to create an interface between the two traditions of epigraphy and manuscript writing. In some cases we can even see the page serving as a proxy monument, both in terms of the manuscript's comparable status as a prominent and enduring written artefact and also of its function in reminding the reader of an earlier, superseded, written tradition. This epigraphical transposition was allowed for in early medieval England due to the existence and continuing circulation of an alternative script-the runic futhorcthat was closely associated with inscription and with the writing traditions of an earlier period. This epigraphical memory, the received knowledge that runes were once engraved on durable objects, meant that the script was able to serve as a marker of an alternative textuality when employed in manuscripts, drawing attention to the physical, tactile, and ornamental aspects of writing and invoking associations with the durability of letters carved in stone or metal. Furthermore, because runes stand out on the manuscript page and were associated with a different form of writing on different media, poets and scribes were able to use this alternative script to highlight the written word itself and draw attention to its unique attributes.

Epigraphy in transition: the runic tradition
The critical consensus is that the runic script originated at some point in the first or second century AD amongst the Germanic tribes of Northern Europe. It was a script almost certainly inspired by contact with either the Roman writing system or-less plausibly-Etruscan or Greek scripts, but it departed from these models in several key respects. Firstly, it seems to have been devised specifically for incising on durable materials due to the lack of curved lines that are difficult to carve. It has also been proposed that a lack of horizontal lines that would be hard to distinguish against the grain of a piece of wood suggests that runes were devised specifically for writing on this medium: though few inscriptions in wood have been found from the early period, it is probably right to assume that this was a common material used for runic inscriptions in contrast to the use of papyrus, parchment, or clay tablets. 1 Indeed, this is a distinction made by the sixth-century Merovingian bishop and letter-writer Venantius Fortunatus, who in one epistle gently mocks a friend who had not sent him news by suggesting that he could try a different script if that was what was holding him back from corresponding, or that in extremis 'Barbara fraxineis pingatur runa tabellis' ('the barbarous runes might be painted on ash-wood tablets'). 2 In contrast to the lengthy correspondence Venantius was hoping to receive from his friend, most of the runic inscriptions preserved from the earliest period are short: this brevity is in part dictated by the materials used and the rather laborious process of engraving or incising the characters. The runic script was also consciously detached from whichever writing system inspired it, including rejecting the 'abc' order of Mediterranean alphabets, devising letter forms without precedent in other script systems, and providing each character with a meaningful name. 3 Runic writing in its earliest phase is particularly well attested in southern Scandinavia, but the tradition stretched across Northern Europe from present day Ukraine to the Netherlands. 4 It was the only writing system known to the Germanic tribes who came to Britain from the Continent, and though there may have been limited influence on the script from, for example, written traditions amongst the Romano-British, it is not until the conversion of Anglo-Saxon England in the seventh century and increasing Scandinavian contact with Christian Europe in the Viking Age that we can really begin to trace the interaction of manuscript and epigraphical traditions with any degree of certainty. After this point, the influence of manuscript writing practices on the runic tradition is fairly easy to discern, at least in the script's younger incarnation in Scandinavia and Early Medieval England (from around the eighth to eleventh centuries) and in its later medieval forms in Scandinavia. 5 One prominent 'transitional' inscription from the Scandinavian tradition is the elaborate rune stone DR 42, raised at the royal seat of Jelling in Denmark. 6 This monument-and the inscription refers explicitly to kubl : þausi or 'these monuments' rather than to 'stones' or 'runes'-was commissioned by Harald Bluetooth in the late tenth century in honour of his father and mother, as well as in praise of his own achievements in winning all of Denmark and (at least parts of) Norway for himself, and making the Danes Christian. In light of this last statement, and the fascinatingly unorthodox depiction of Christ's crucifixion that it accompanies, this monument is commonly referred to as the country's 'baptismal certificate'. 7 As anachronistic as this may sound, it is perhaps not too far off the mark to label this monument a 'certificate' or 'document', as the inscription is clearly and deliberately borrowing the iconography of the book, which was closely associated in this context with the Christian mission. This is particularly clear on Side A of the stone (Fig. 1), which gestures towards the conventions of manuscript layout, including a carefully ruled runic text and an elaborate interlace design in the left-hand margin that resembles an illuminated initial. This represents a borrowing so striking that Else Rosedahl dubs the monument 'en bog af sten' ('a book of stone') (Rosedahl 1999: 235-44). Harald Bluetooth's monument was clearly engraved by an accomplished rune carver who was well versed in the Scandinavian runic tradition: the runes are well made, and there are few mistakes barring the slightly inconsistent use of punctuation marks between words. Yet this same rune carver was also an innovator, helping to establish the Mam-men style of art which flourished in southern Scandinavia in the mid-to late tenth century, and of which Jelling II is 'the most important representative' (Wilson 2008: 330).
There is some debate about whether the intricate artwork on the rune stone, including the lion of Side B, represents purely an evolution of native Scandinavian traditions, or whether there are influences from, for example, the Anglo-Scandinavian traditions of the Danelaw or epigraphy in the neighbouring Ottonian empire (Wilson 2008: 330). The same difficulty exists in linking the layout of side A of the stone with any particular manuscript or manuscript tradition: the rune carver may have been exposed to books at Jelling, within the newly evangelized Danish royal court, or in England, or have been influenced by books produced in the empire of Otto I to the south. As the illumination from a late tenth-century Ottonian manuscript below serves to illustrate (Fig. 2), the layout of the runic text clearly mirrors that of deluxe pages in contemporary manuscripts, including the wrapping of the writing space to the top and left with an elaborate illumination, and the wide ruling between short lines of text. This talented rune carver consciously adapted their repertoire to incorporate these most obvious visual features of the manuscript page whilst continuing to follow the conventions of runic epigraphy in the use of the script itself.
The impression that this was a conscious borrowing of manuscript aesthetics to make a point about the newly Christian regime at Jelling is reinforced when we set DR 42 alongside the older-and more modest-Jelling I stone (DR 41) commissioned by Harald's father, Gorm the Old, which is engraved in vertical columns as if mimicking a series of runic staves (Fig. 1). 8 In the carving of the Jelling II stone a transposition has clearly taken place between the runic epigraphical tradition and the book writing conventions of layout and text presentation introduced with the early Scandinavian church. This is not the first example of transposition that could be traced, of course, but it is a striking illustration of the meeting of traditions and the tangible ways in which manuscript practices could influence memorial inscription.

Transposition in Anglo-Saxon epigraphy
In the runic tradition of Early Medieval England, the impact of both 'book writing' and cross fertilisation from the Latin epigraphical tradition is even more pronounced, as evidenced by the numerous mixed Roman and runic inscriptions, and runic epigraphy arising from a clearly learned Christian milieu at a relatively early date. 9 Both these elements are visible on perhaps the most famous of all runic artefacts from England, the eighth-century Franks Casket, which represents a deliberate fusion (and literal juxtaposition) of Germanic, Roman, and Christian traditions. Included in the casket's design are depictions of the tale of Romulus and Remus, the first Roman Jewish war, the feats of a character named AEgili, and the legend of an obscure Hos and Ertae, all bordered by runic inscriptions referring to the scenes depicted. On the front panel of the casket a mirrored diptych brings two very different stories into dialogue, that of the adoration of the Magi and the legend of Weland the Smith (see Fig. 3). These two scenes represent central narratives from the Christian and Germanic legendary traditions, and their positioning suggests that the reader was expected to recognize and relate one story to the other. Of course, another important aspect of this synthesis of traditions is the use of both runic characters and Roman letters on the same object, with the runes carved in relief and contributing in quite a striking way to the decoration of this beautiful object, possibly intended to hold a holy relic. Indeed, on the front panel of the casket (Fig. 3), the large runic inscription surrounding the scene refers not to the legends portrayed, but to the materiality of the object itself and the fate of the 'terrible king' (or whale) whose bone was used to make the casket after it was beached on the shore. The runes are thus closely associated not only with the durable medium of the whalebone casket, but also with the back-story of how this material came into the artist's hands.
The other runic legends do refer to the scenes depicted, but that does not necessarily mean that the script is being used without thought for its wider symbolism. Indeed, in certain places on the Casket the runic script is probably being employed to highlight both Old Testament and early Germanic written and cultural traditions referred to in the imagery in contrast to the association of Latin with the New Testament (Birkett 2017a: 17-21). These older traditions include an obscure scene on the right panel, involving the otherwise unknown figures Hos and Ertae, referred to in the accompanying runic legend. In this particular runic inscription non-runic symbols are employed, as well as several that resemble older forms of the runes, perhaps consistent with the perceived venerability of the Germanic legend depicted on this panel. Yet even here, where the carver is attempting to evoke a pre-Christian legend using a deliberately obscure writing system tradition, the crossover with the manuscript tradition can also be discerned. The cryptic runes appear as part of a vowel substitution code, by which invented characters or a system of punctuation is used to stand in for the vowels in a text and thus to hide or delay apprehension of its meaning. This was a conceit which was fairly common in monastic contexts and 'particularly popular amongst Anglo-Saxon scholars since the eighth century' as Parsons points out (1998: 443). The influence of the scriptorium on the epigraphy of the Franks Casket is perhaps not that surprising in light of what Webster and Backhouse refer to as the 'ostentatiously erudite' and culturally syncretizing context in which this object was created, almost certainly a Northumbrian monastic community (1991: 103).
Even the Latin text that appears on the rear panel (written partly in Roman script) turns away from Latin epigraphy and towards the Latin minuscule of the scriptorium (Fig. 4). This panel depicts the first Jewish-Roman war, and part of the inscription reads Hic fugiant Hierusalim habitatores ('Here the inhabitants flee Jerusalem') with the final word 'inhabitants' rendered in runes in contrast to the Roman script used for the rest of the statement. 10 Curiously, the Roman script used seems to be 'imitating the pen strokes used to form letters in Anglo-Saxon manuscripts': it is deliberately turning away from the angular forms of the runic script and towards book writing, and is thus consciously situated at the interface of epigraphical and manuscript traditions (Webster 2012: 49 note 41). The effect of this switching of scripts, as well as the coding of letters on the right panel, and retrograde writing on the front (another feature of early runic epigraphy), is to break the fluent reading process-even for someone familiar with runes-and to refocus the attention on the form of the letters, some Roman, some runic, and some invented. Writing on the casket is thus not an inert graphical representation of spoken language, but a symbol in itself of different written traditions and the various ways that information can be coded and displayed.
As the extraordinary textuality and materiality of the Franks Casket makes clear, the issue of transposition between manuscript and epigraphical practice is more complex and far-reaching than the formal transfer of layout or text convention: scripts or writing practices may be deployed for particular effects, and this transposition involves not only different epigraphical traditions (runic and non-runic), but on occasion also different languages and sources of culture. 11 There are several other examples of monumental texts at the interface between traditions that could be discussed at this point, including St Cuthbert's coffin, with its mixed Roman and runic tituli-labels naming Christ, the apostles, and other biblical figures-engraved into the wood, indicating, perhaps, a community that used these scripts interchangeably. The famous Ruthwell and Bewcastle Crosses also have Roman and runic used on the same monument, in the former case recording an Old English poem which situates the cross in a heroic pose as it participates in the crucifixion of Christ, the incised runic characters vividly mirroring the cross's statement that it was '[m]iþ s[t]re[l]um giwundad' 'wounded with arrows'. 12 As even these few examples make clear, there was no sense in which runes were deemed inappropriate for Christian monumental inscription or for memorializing contexts, even if different material and cultural associations may have been brought to bear in the use of the two different scripts, one rooted more firmly in the manuscript tradition and the other gesturing towards a vernacular epigraphical tradition. In fact, Parsons has put forward a compelling argument for a conscious reorganisation or reform of the Anglo-Saxon futhorc within the Church in the second half of the seventh century (Parsons 1999), 13 meaning that a learned process of standardisation may form the basis of virtually all runic inscriptions that follow. This includes inscriptions from the informal context of a personal prayer to 'Almighty God' on the Whitby Bone comb inscription, which was found in a rubbish pit, to the monumental runes of the Hackness cross, incorporating another ostensibly 'learned' practice of cipher runes. This is a practice attested by the so-called ísrunar tract-a short collection of different runic ciphers recorded in manuscripts of Anglo-Saxon provenance circulating on the Continent and probably associated with eighth-century missionary activity from England-and thus possibly also a conceit dreamed up by ecclesiasts (see Derolez 1983: 83-9). In other words, it is hard to conceive of an English inscription from the eighth century onwards that is not influenced in at least some way by book-learning and writing practices within the scriptorium.

From monument to page
In light of the early ecclesiastical uptake of (and influence on) runic epigraphy in England, it is quite surprising that the division between the runic tradition proper (as it is referred to by many runologists) and the tradition of runica manuscripta-the manuscript record of runes-is still so pronounced in runic scholarship. 14 The prevailing view is that runic epigraphy is an early English interpretation of a common Germanic tradition and thus primary, and manuscript runes are an antiquarian response to a tradition already in decline, a preservation and reuse of the script as a plaything of scholars, riddlers, and 'the immature who delighted in strange and cryptic alphabets' as Page memorably puts it (1999: 42). This perceived division, and the 'romantic view of the epigraphical tradition' that Parsons identifies in Page's lifelong characterisation of runica manuscripta (1994: 197) has almost certainly led to a downplaying of the influence of manuscript practices on the examples of runic epigraphy just described, even though Page himself admitted to reconsidering his position and being 'more ready to see links between the two' when he reflected on the 'way forward' for scholarship on the Anglo-Saxon runic tradition (1991: 19). Moreover, this prevailing attitude has led to an almost total disregard of influences in the other direction: namely the way monumental runic practices and epigraphical conventions may have found their way into the pages of the book. Parsons argues, convincingly, that 'it is difficult to imagine how two separate runic traditions could have been maintained, one indoors and the other out ' (1994: 197). The possibility of epigraphical transposition-the borrowing of conventions associated with inscription for specific textual effects reliant on the epigraphical tradition to which they refer-will be explored in the second part of this article, and a case will be made for a similarly meaningful deployment of runes in manuscripts to that witnessed in the epigraphical tradition encompassing the Franks Casket.
In his seminal work Runica manuscripta: the English tradition, Derolez divides manuscript runes into several categories, including futhorcs circulated in the context of other obscure or invented alphabets, treatises on runes such as the ísrunar tract mentioned previously, alphabets in which the traditional futhorc order is rearranged to correspond with the Latin alphabet, the circulation of runes within the Inventione Litterarum tradition, and finally non-alphabetic uses of runes-into which category the author places scribal signatures, runic abbreviations, notes in runes, riddle clues, and decorative uses of the script (1954: 404). Derolez gives these last uses of the script short shrift, the implication being that scribes are casually and uncritically making use of the circulating runic material in order to save space, write in code, and satisfy a monkish fascination with enigmas and esoteric knowledge. However, these occasional uses of runes are precisely where we see the script employed for its effect rather than simply as recorded alphabet lore, and where we can identify certain associations with monumental writing being brought to bear in a manuscript context.

Runes as a display script
Perhaps the most obvious category of runica manuscripta to begin with is the use of the runic script solely as a decorative or display script. Such uses are found most often in manuscripts produced on the Continent in centres closely associated with Anglo-Saxon missionary activity in the eighth and ninth centuries. In fact, a large proportion of the surviving runica manuscripta arise from amongst this diaspora, and there seems to have been an overall decline in the accuracy of the rune lore preserved as links with England dwindled. Derolez is right, in this context, to talk about the 'bookish character of the Continental runica manuscripta' and to view it as a tradition of the scriptorium (1954: 428), perhaps preserved out of a lingering sense that runes were part of the heritage of these ecclesiastical communities. However, despite their dislocation from the practice of inscribing runes, there is a chronological overlap with the epigraphical tradition in England as Parsons points out (1994: 198), and on occasion a tentative link can perhaps still be discerned, particularly in the recourse to the futhorc as a display script.
One such instance of runes used in manuscripts is a scribal signature described by Derolez as 'a large ornamental inscription in runes, alternately one line black and one red ' (1954: 404), and found in a ninth century manuscript originating from Reims in France. 15 The display colophon names the scribe (possibly a certain Josephus) and is more elaborate than most scribal signatures, though conventional in its offering of the copying of the book as an act of devotion. As is clear from the large illuminated initial that occurs on the facing page of this manuscript, the runes certainly did not have the monopoly as a display script in this manuscript, but for the scribe who had finished his stint at fol. 6v, runes were deemed the appropriate script for an ornamental colophon.
Considering the indulgent use of space, the lack of punctuation between words and the sheer visual impact of the black and red letters, it is clear that this is a script used for its display qualities rather than for ease of communication, and I think Derolez is right to refer to this as an 'inscription'. It also echoes several key features of monumental epigraphy; namely the visually striking use of red and black letters, the liberal use of space in comparison with the body text, and the fact it is memorializing the individual responsible for the creation of the text-object.
The use here of rubricated runes, or runes rendered in red ink, itself provides a link with the epigraphical tradition: whilst only one or two examples of Anglo-Saxon epigraphy using paint to highlight the letters have survived (hardly surprising given their exposure to the elements), 16 we can be fairly sure that monumental inscriptions such as those on the stone crosses were once brightly painted: this is based on the recovery of fragments of Scandinavian rune stones (often embedded in the fabric of later buildings) with traces of paint as well as comparison with other surviving examples of early English art. Prag goes as far as to suggests that there is 'overwhelming evidence' that monuments such as the Ruthwell Cross and Bewcastle Cross were coloured, with contemporary fresco painting and manuscript illumination both informing Manchester Museum's painting of a cast of the monument in 1999-2000 (2010: 1). 17 The Josephus scribe, operating in a scriptorium in northern France, almost certainly had no direct experience of runic monuments raised in England at a much earlier date, painted or otherwise, and we cannot talk of 'borrowing' epigraphical techniques here even if this diasporic community still understood that runes were once associated with monumental writing. However, there is an interface of kinds based on the intended effect of the runes and the rationale for their use.
Even if once picked out in red paint, the runes on the Ruthwell Cross are hard to read even for someone well versed in the script, and linguistic communication is secondary to the display quality of the runes and the erudite visual pun of engraving the reddened and 'wounded' cross with its own history. For Josephus also, rubricated runes are not chosen for ease of communication: it is their display qualities that must have appealed, along with a recognition that this angular script was not a book script, but a script that, like the Roman square capitals that preceded the more rounded Uncial book hands or the majuscules used for Latin inscriptions, belonged to a world of epigraphy. In other words, Josephus is borrowing a display textuality of rubricated letters and an exotic angular script in order to create a striking colophon in a manuscript. He is, I would argue, fashioning the page as a monument to his labour; in effect 'setting his name in stone' through the decorative runes and memorializing his achievements using a script valued more for its display qualities and epigraphical associations than its utility.
Another 'very puzzling inscription' drawing on the display qualities of the script is found in a manuscript written in the late eleventh century, possibly at Hastière in Belgium (Fig. 5). Here a striking illustration of Christ's Descent from the Cross includes a variation on the titulus 'Iesus Nazarenus Rex Iudaeorum' transliterated into Greek capitals, Roman minuscules, and runic letters and written onto the cross head (Derolez 1954: 414-15). In this context, we have a true interface between a monumental inscription and the manuscript page, with the three scripts used to 'engrave' the image of the cross-head. The runes, labelled with a B for Barbarice ('vernacular') following the G and L denoting Greek and Latin, are presumably intended to stand in for Hebrew letters as a similarly 'exotic' epigraphical script: once again, they have been chosen to represent an inscription on this illustration of the Deposition of Christ primarily due to their appearance.  6). Directly above this carpet panel (which brings to mind the vine-work motifs of monuments such as the Ruthwell Cross) we find a series of fairly large and prominent runic characters, which were incorrectly glossed at a later date by an individual who assumed they were written in the same order as the alphabet. In fact, the runes are in a sequence that suggests they may represent a meaningful inscription-perhaps using a transposition code or another means of obscuring the message. Page identifies the runes as being written in a tenth-century hand, but admits to having 'no idea what their significance is ' (1999: 212), which is perhaps a little defeatist: even if the linguistic meaning of the sequence remains unclear, the context can tell us something about the symbolism of the runes and the reason they are being used on this flyleaf. In total, there are six distinct images drawn on this manuscript page, all but one with a small label in Roman script (see Fig. 6). The runes, however, are not being deployed as labels, but as a display inscription to which Roman minuscules are added as another explanatory gloss.
The effect of these decorative runes is similar to the vine-work panel: both are visually striking and intended to be viewed primarily as images, and this is exactly how the medieval glossator viewed this sequence when he labelled it as an alphabet sequence rather than a meaningful text. The same is true in this manuscript context: the runic sequence is writing intended for ornament and display, gesturing towards 'monumental' inscription even though written on vellum. Of course, the display quality of runes is also exploited in literary contexts, where their epigraphical associations become part of the meaning of the text in which they appear. Often a runic conceit relies on the fact that the script is visually distinct from the surrounding Roman letters, and it is often made more visually striking by increasing the size of the characters, using thicker pen strokes, including a prominent punctus before and after each character, or including serifs on the letters. This is true of the famous runic riddles in the Exeter book, which employ runes as clues-sometimes anagrams-that need to be pieced together to solve the riddle. 19 Here the unusual, angular appearance of the runes helps to draw attention to the riddle within a riddle, with embedded anagrams in particular representing as much a visual puzzle as a linguistic one. Sometimes, as in The Husband's Message (a poem based around the conceit of a portable runic inscription relaying its message to a loved one), the use of runes at the close of the poem represents an embedding in the manuscript of a different kind of textuality, one deliberately mimicking the carved inscription. Indeed, as with the Orosius flyleaf inscription discussed above, the linguistic meaning of this embedded runic message is still much debated, and the hermeneutics of the runes seem to rely once again on their associations with a different kind of literacy transposed from the world of inscription: perhaps not so much a 'transliteracy of obsolescence' as Foys (2012: 143) has it, or a gesture towards a dead tradition, but a form of writing anchored more firmly in epigraphy. The runes used in these contexts signify carved and coded messages, the tactile medium of engraved wood and stone, and the visual impact of monumental inscription, even as their linguistic message remains obscure.
Perhaps the best example of this borrowing of what I have called elsewhere the 'ornamental textuality' of runes occurs in the curious poem Solomon and Saturn. 20 This complex poem involves a dialogue, or wisdom contest, held between the biblical patriarch Solomon and his pagan interlocutor Saturn, and it ranges widely through different fields of esoteric knowledge arising from 'a bookish circle interested and informed in the tradition of learned riddling' (Anlezark 2007: 104). It is preserved in two English manuscripts, two poetic dialogues separated by a prose sequence, in Cambridge, Corpus Christi College MS 422, and a fragment (fortuitously containing the opening section missing in MS 422) squeezed into the margins of the Old English Bede (Cambridge, Corpus Christi College MS 41). The first poetic dialogue is almost exclusively concerned with Solomon explaining the power of the written pater noster prayer to achieve all kinds of desirable effects including victory in battle, escape from spiritual constraints and the very literal crushing of the forces of evil. It is remarkable for letters to be given the hyper-physicality they have in this poem: the anthropomorphised characters dash the faces of devils against the rock, scattering their teeth; they flay and bludgeon and smite, and they are described as 'stronger than the foundation of all stones' (l. 76). 21 It is perhaps no surprise that this extraordinary expression of the materiality of the word is accompanied by the use of runes to represent the individual characters of the pater noster in Corpus Christi MS 422, not as an afterthought, nor as 'a symptom of transmission in circles interested in alternative alphabetical systems', as Anlezark has it (2009: 29), but as the letters that precede their Roman equivalents and that act as the primary visual referent. We are told that the runic letter ᛏ, t, injures the devil, 'stabbing him in the tongue, twisting his throat and shattering his jaw' (ll. 94-95) whilst the letter ᛋ, s, will 'propel [the fiend's] cheek forward against strong stone, and strew teeth around the host of hell' (ll. 113b-115a).
For letters imagined to have a tangible presence, a physicality of stone and steel and broken teeth, it seems the runic script-with its lingering association with monumental epigraphy, with incision into stone, and with engravings on weapons-was the natural script to use. 22 These are runes whose 'function seems to be purely ornamental', but not in the reductive way that Derolez implies: rather, through their ornamental qualities they represent a striking manifestation of the physical, visual, and material presence of the written word (1954: 420). Indeed, when the poet refers to the curious image of 'the palm-twigged pater noster' (l. 39) in this same poem, he may be thinking of the prayer written in runes: indeed, we have several carved runic pater noster prayers surviving from the medieval Scandinavian runic tradition, including a complete prayer carved on a lead table from Ulstad, Norway. 23 Etched into stone or metal with sharp instruments, runes had a physical presence not as readily apparent in the book scripts Anglo-Saxon scribes would have been using habitually: and it is this physical, tactile association that is surely being drawn upon here in the anthropomorphism of the letters. To return to the title of this article, these hyper-physical letters represent a materiality that has been transposed from the domain of epigraphy to the manuscript page for a particular effect: reminding the reader that letters can have a physical presence in the world.

Runes as maker's marks
Further evidence of a close association between runes and ornamentation can be discerned in the category of scribal signatures. As Symons points out, many scribal colophons 'incorporate considerations for their visual appearance', and the use of runes might be thought of as just one strategy (along with rubrication, layout, or even acrostics) for ensuring that the signature stood out from the page and for highlighting both the craftsmanship of the scribe and 'the material construction of the book' they contributed to (2016: 105). Runes are, however, used regularly enough amongst the Alcuinian diaspora on the Continent in the late eighth and ninth centuries that Warwick Frese felt confident to pronounce 'a rage for the insertion of runic inscription into their manuscripts ' (1975: 312). These include a scribe from St. Gall in Switzerland called Ratger (Fig. 7) who signed off his copying stint in runes, an individual from St. Amand in Northern France called Agambertus, a certain Cundpato from Freising, and another Madalfrid from the same community, as well as the Josephus previously discussed. 24 In each of these cases, the scribe demonstrates an interest in obscure writing systems; 22 A sword engraved with bealwe bocstafas or 'baleful letters' is referenced in this same poem at lines 161-3, and the poet's 'technical knowledge of metalworking' has been documented (Anlezark 2009: 11). 23 Designated N 53 (https://www.runesdb.eu/en/find-list/d/fa/q////6/f/2052). For a survey of the pater noster inscriptions, see Zilmer (2013: 129-71). See also the more complete discussion of the Solomon and Saturn conceit in Birkett (2017a: 88-91). 24 For a more extensive account of these continental colophons, see Derolez (1954: 403-16  indeed Derolez describes Agambertus as having 'a special liking for strange alphabets', his signature colophon amounting to 'a regular display of cryptic systems ' (1954: 406). The other scribes mentioned dabble in Greek and several invented alphabets, including the notae sancti Bonifatii popular in Anglo-Saxon circles (Gameson 2002: 26), yet when it came to signing their names, all five scribes mentioned above (having several exotic alphabets at their disposal) turned specifically to runes. It is conceivable that this was because, if these scribes were indeed part of the Alcuinian diaspora, they associated runes with an Anglo-Saxon heritage. But we should also consider whether there is a further influence at play here that may have been transmitted along with knowledge of runic characters: namely, an association with engraving, memorialisation / commemoration, and even 'maker's marks', perhaps persisting as an oral gloss to the circulating rune lore. An understanding, in other words, that runes were an epigraphical script particularly suited to recording a name for posterity, and that by using runes you were in some way setting your name in stone.
Such an association is perhaps even clearer in the sixth-century manuscript supposedly owned by the eighth-century Anglo-Saxon apostle to the Germans, St. Boniface, and glossed in his hand. This manuscript includes a runic inscription on its cover, representing a slightly garbled threefold invocation of IOSEPI (Derolez 1954: 414), a reading that makes sense as the manuscript contains a copy of the so-called Pseudo-Hegesippus, a Latin translation of Flavius Josephus' The Jewish War attributed to Saint Hegesippus. 25 The use of runes to attribute authorship in this way, etched into the cover, is remarkable, particularly as runes only feature here and on the back cover (in the form of an alphabet): the script does not feature within the pages of the manuscript itself. Perhaps what we have here is a preserving of a distinction between script uses: Roman as a book script, both for the sixth-century body text and its eighth-century notes and corrections, and runes as a script for incising the cover and naming the author.
The IOSEPI inscription may have been made whilst the Kassel manuscript was in Northumbria or once it had already made its journey to the Continent, but a similar drypoint can be more firmly connected with England. Page notes an inscription in the St. Petersburg Gospels, dating to the late eighth century, and probably made in Northumbria, rendered as eþe=lst=dryþ. Here the runes are deployed in a fairly elaborate manner (including two ligatures) to record what Page presumes to be two names- Eþelstan and Eþeldryþ (1999: 198). In both these cases we are dealing with drypoint etched into the vellum with a sharp instrument, which is in itself a mode of writing transitional between epigraphy and pen and ink, further evidence, perhaps, that the recording of a maker or artificer's name in runes was associated with a textual world outside the book. 26 If we look at Cynewulf 's runic signatures-perhaps the most famous example of Derolez's occasional uses of runes-we can see that this early English poet may also be drawing on an association of the script with epigraphy and with the tradition of naming the maker or the artificer. Halsall suggests that there may have been a strong association between runes and memorializing inscriptions-such as the Hackness Cross or those found recently at Lindisfarne-and that to a certain extent Cynewulf was taking his inspiration to 'remember him by name' from the tradition of engraving a name on stone (1989: 477-86). 27 Warwick Frese in turn suggests he may be looking back to an even older tradition, namely the 'centuries old custom… to runically engrave the signature of the artificer', as attested in the Anglo-Saxon tradition most memorably by the short inscription 'Ludda repaired the brooch ' (1975: 313). 28 There is undoubtedly more going on in Cynewulf 's theologically engaged and hermeneutically significant signatures than a simple transposition of 'permanence', 'legacy' or 'mark of ownership' from the world of epigraphy to the manuscript page, 29 but he is surely bringing elements of this epigraphical tradition to bear in his use of runes, incising a name within the poem as an act of memorialization as well as devotion.

Runes as a script of authority
Runes tend to serve a very niche function in early English manuscript culture, and it is rare for a runic note to intervene in matters of textual or literary authority. One exception to this tendency is a correction to a scribal emendation in an eleventh-century manuscript of the Old English Bede, a seminal text of ecclesiastical history which every learned individual could be expected to be familiar with (see Fig. 8 It is notable that the eleventh century is a late context for runic writing, even for the writing of runes in manuscripts. Some more evidence is emerging that the epigraphical tradition persisted in certain areas for longer that we once realised-perhaps even into the eleventh century. 31 This is certainly the case in the south of England where this manuscript was written down, and in the collections of Exeter Cathedral where it may have been glossed and emended. In these contexts, the use of runes represents the revival of a script that had not been used for monumental epigraphy for centuries. To answer the question of why the scribe turned to runes to write the note, we have to think about the lingering associations that the script may have carried, even as the epigraphical tradition became a thing of the past in England. The manuscript emendation concerns a figure called Baldhild, whose name has been changed to Brunnhild by a second scribe. This Frankish Queen was, according to Bede, supposed to have persecuted and killed the Bishop of Lyon, and presumably her name was changed in light of the fact that Baldhild was later canonized, somewhat awkwardly considering the tyrannical behaviour that Bede recounts. The runic glossator seems, however, to have been uneasy about this change to the text and the challenge to Bede's authority, and the runic note reads 'xii and xxx more', which may refer to the discrepancy in dates between the reign of these two queens, though it has also been suggested that it relates to a land grant mentioned at the end of this manuscript page (see Birkett 2012: 469-70;Page 1999: 198). So why would the glossator turn to runes for either of these purposes? Was he using an obscure text to hide his comment? This would rather defeat the point of making the correction, unless it was for personal satisfaction alone. More plausible is that the corrector turned to a script that he believed had authority, both because of its association with a past age, and perhaps in the mind of an eleventh-century scribe, with the Northumbrian locale of Bede. More importantly, it suggests a lingering association of runes with epigraphy and with the permanence that comes from engraving into durable media. In a recent article, Findell and Kopár suggest 'a special prestige [of runes] in a commemorative context and an association with social memory and the ancestral past' (2017: 133), an association that I suggest extends in certain cases to the manuscript tradition. This note is being placed in runes at this late date not as a riddle, but as a statement of enduring written authority in support of the venerable Bede. It is an inscription on the page, borrowing a different textuality-older, epigraphical, monumental, authoritativeto make a point within the manuscript environment. This is an antiquarian use of runes, maybe, but with awareness of the epigraphical tradition that was being transposed into the pages of the manuscript book.
This kind of textual intervention in runes is unusual, but there are other examples. One of these is a comment in the famous compilation of Old English poetry known as the Exeter Book. In this particular manuscript runes are used fairly extensively, as embedded clues to half a dozen riddles, in the scribal colophons to the two Cynewulf poems recorded in this manuscript (where the runes have to be reassembled to locate the name of the poet behind the conceit) and as abbreviations on a couple of occasions. There are also several clues to individual riddle poems, which use runes to indicate the first letter of the solution. Some of these are inscribed in drypoint, which is of course in itself something of an interface between epigraphy and book writing, relying as it does on the materiality of the page and incising into the vellum rather than overlaying with ink. There is also a slightly longer inscription in drypoint runes beside one of the riddles, a complex little enigma which centres on a man riding a horse with a hawk in hand, and which uses runic characters at key points within the riddle to stand in for their names. Page suggested to Williamson in personal correspondence that the accompanying drypoint inscription may stand for Beo unreþe (be merciful)-with vowels removed in a similar convention to that on the Franks Casket: Williamson takes this as 'a frustrated scribe's (and jokester's) comment on the absurd difficulty of the runic riddle ' (1977: 327). Whatever its true meaning, it is interesting that the reader decided to inscribe his comment in drypoint, rather than using ink, and that he used the runic script for this purpose rather than Roman script. It could be argued that all the reader had to hand was a stylus rather than pen, but if so, why not engrave using familiar Roman letters? I would suggest that this is another tantalising hint that runes were-at this late tenth-century date-still being associated with a different type of writing practice, one that was more rooted in engraving with a sharp object on durable materials.
All the runic interventions discussed so far have involved the transposition of runes onto the manuscript page, but we also have evidence surviving from two poems within this same Exeter manuscript that describe the writing of the letters. One of these is a curious riddle about a copulating pair of animals-Riddle 42 in Krapp and Van Kirk Dobbie's numbering (1936)-in which the narrator describes how he may write on the floor in runes the names of these two brazen creatures: a cock and hen. 32 Again, this is an ephemeral writing practice, but one grounded in inscription, even if the use of a stick in the dirt is not exactly monumental epigraphy, it is still part of that transitional zone between carved objects and book writing. Secondly, we have a depiction of runic correspondence in The Husband's Message discussed at an earlier point in this article, which depicts a speaking object telling the husband's beloved about its journey across the sea and the intentions of its owner, before recounting the message itself in runes. It is interesting that the message is conceived of as being carved on wood, specifically a beam, sometimes taken to be the mast of the ship of which it travels as a public statement of the husband's fealty (see Niles 2003: 204-6), but perhaps more plausibly a type of rúnakefli (rune stick) (Birkett 2017a: 68-9) or a writing tablet (Symons 2016: 75) that has undergone the journey. We have very little material evidence of Anglo-Saxon carving of the kind of wooden slips found in medieval Bergen, but this poem hints again at a practice of inscription in runes that would not have left a material record and that may have continued after the decline of the monumental epigraphical tradition. As a support to this, a late-eighth-century whale-bone writing tablet from Blythburgh in Sussex was found to have runes lightly engraved on the back plate, possibly representing 'trial forms', or occurring as a result of cutting through the wax layer that covered this kind of tablet and that allowed for its scraping clean and reuse (Webster and Backhouse 1991: 81). This and other portable runic objects discovered in recent years strongly suggest that in certain Anglo-Saxon religious contexts, runic inscriptions were more common than the monumental record suggests, perhaps most often carved in ephemeral contexts such as on wood, wax tablets, or-as in the riddle-written in even more transient forms, to be swept clean from the floor. 33 The tantalising evidence of The Husband's Message and the Exeter Book riddles takes us back, in a roundabout way, to the beginning of this article, and the proviso that whilst the script seems to have been devised for engraving onto wood, almost no evidence survives from the earliest period for this kind of inscription. We have to be very careful with negative evidence of this kind, but as this survey of runica manuscripta suggests, such a strong link between runes and the materials they were written on may have loomed large and influenced the way that the script was used and deployed within the manuscript page, often in full awareness of the letters' physical presence, material associations, ornamental qualities, and even, perhaps, their monumental legacy. For this reason, I think it is wise not to draw too sharp a distinction between the worlds of epigraphy and manuscript practice, and to acknowledge that influence almost certainly goes in both directions, even if in late Anglo-Saxon England, this influence came in the form of what I have called 'epigraphical memory' rather than direct contact with a living tradition of monumental inscription in runes.

Conclusions
There is clearly a lot of scope for reconsidering the transfer of manuscript practices to the epigraphical runic tradition, with the examples of the Jelling II stone and the Franks Casket representing two of many inscriptions that might be described as transitional, influenced by the layout, aesthetics and prestige of manuscripts (as in the case of Jelling II), or by scribal conceits such as ciphers and learned alphabet lore (in the case of the Franks Casket). Whilst such influences have long been recognised by runologists, borrowing in the opposite direction-or the impact of monumental inscription and epigraphical memory on the use of runes in manuscripts-has received minimal attention. Whilst the examples gathered here do not amount to anything so cohesive as a defined movement or even common understanding of what this script represents, they do hint at a continuing association of runes with epigraphy and with a different kind of writing rooted in memorialisation, artifice, and display.
Even if the page does not always become a monument through the use of runes, as it seems to do in the case of Josephus' colophon, the Orosius flyleaf, and Cynewulf 's signature conceits, we should be open to the possibility that certain associations or conventions were borrowed from the world of epigraphy and used for particular effects in Anglo-Saxon manuscripts, whether that be for memorialising a name in the pages of a book, claiming an authority commensurate with an earlier writing tradition, or illustrating the literary depiction of letters as physical beings stronger than the foundation of the stones into which they are inlaid. Runes are minor players in the development of a distinctly Anglo-Saxon textuality, but the interface between runic epigraphy and runic interventions in the world of the book should not be ignored, not least because it encompasses some of the most profound early English meditations on the written word and its physical presence in the world.