“Whistle While You Work”: Scribal Engagement with Old English Poetic Texts

I t is well known that Anglo-Saxon textual reception was mostly not a silent or individual experience, but a communal and auditory one. There is considerable evidence that scribes understood this reception context, and that some manuscripts were “deliberately designed to assist the delivery of acoustic texts”. It is also clearly understood that some texts were designed to be seen as well as to be heard: appearance could shape meaning. Hearing and sight often worked together; in Katherine O’Brien O’Keeffe’s terms, “interpretative information appears in graphic directions determined and mediated by the

subjectivity characteristic of oral performance". 3 In the following discussion, I consider not the sensory experience of receiving and performing texts, but the sensory experience of producing them. It is my contention that Anglo-Saxon scribes could and did engage with Old English poetic texts. This in turn implies a scriptorium where scribes did not have silently to process the individual graphemes of the texts they reproduced, but where they could hear and, to some degree, represent metrical rhythms in their writing. 4 There are four major codices of Old English poetry, all produced within a relatively short period of time from the mid tenth to the early eleventh century. They are the Vercelli Book, the Exeter Book, Junius 11, and the Nowell Codex. 5 All four contain only Old English, though they have few other similarities.
As some examples of their differences, out of the four the Vercelli Book and the Nowell Codex include prose alongside poetry; the Nowell Codex and the Exeter Book include both religious and secular pieces; the Exeter Book and the Vercelli Book have just one scribe; Junius 11 and the Nowell Codex are partially illustrated. Comparing scribal practice in each, then, is not straightforward; the challenges becomes more marked given variance between texts within individual manuscripts. As a very general rule "scribal practice has to be considered poem by poem". 6

Scribing as Sensory, Spiritual, and Mechanical
There can be no doubt that scribing was and is a sensory experience. Cheryl Jacobsen points out that palaeographers tend to describe scribal marks in terms of their appearance, where a scribe describes marks in terms of how they are made. So, for instance, "[w]hat the scribe calls a down-stroke the palaeographer calls a minim". 7 Even the sensory experiences of scribal tools can be sharply imagined. In the frontispiece to an early eleveventh-century manuscript from Saint-Vaast, the scribe says: Cum librum scribo, Vedastus ab aethere summo Respicit e caelis, notat et quot grammata nostris Depingam calamis, quot aretur pagina sulcis, Quot folium punctis hinc hinc laceretur acutis.
When I write a book, St Vedastus looks down From highest heaven and notes how many letters I depict With my pen, by how many furrows the page is ploughed, By how many sharp points the folio is punctured here and here. 8 Such keen awareness is bound up with spirituality: in the process of wounding, the flesh of the folio starts to sound almost like the tortured flesh of Christ, an association which becomes stronger when Vedastus goes on to promise remission of sins to the same number as the letters, points, and furrows. The connection is spelt out in an early twelfth century addition to a miscellany from Durham, where each tool has its own allegorical association. For instance: The chalk with whose fine particles it [the parchment] is whitened signifies an unbroken meditation of holy thoughts, which makes our conscience resplendent. 9 Further, scribes could use their presence in the process of transmission to insert themselves into the commune that the copied texts created between readers and heaven. This applies to scribes working on volumes in the vernacular, and containing Old English poetry, just as it does to those working on Latinate and scriptural productions. The Nowell Codex ends with Beowulf and Judith but opens with three prose texts, the first of which is a version of the Passion of Saint Christopher. There, the saint's final prayer reads: drihten min god sy ÿle gode mede þam þe mine þro ungaÞarite 7 þa ecean edlean þam þe hie mid tearum raede . 7 Lord my God, give a good reward to any who write about my suffering, and eternal reward to any who read it with tears. 10 This mention of the scribe does not appear in all versions: it is not 'original' and was presumably introduced by a scribe at some phase of the text's reproduction, 11 as a more thoroughly embedded version of numerous scribal colophons which invite prayers for the scribe. 12 To some extent, as Fred Robinson argues, scribes could claim equivalence with authors in terms of spiritual significance, both "intermediaries through whom God speaks". Anglo-Saxon illustrations of historical scribes emphasise this spiritual aspect of their work. 14 One example is that of St Matthew in the Lindisfarne Gospels, where the saint as scribe is shown sitting at ease, writing in his lap in relatively luxurious surroundings. 15 A haloed figure, half-hidden behind a cur-tain and holding a book, watches the evangelist, presumably figuring the divine word of God and true Book for which Matthew is an intermediary. Above the saint's head, his symbol of a winged man blows a trumpet in an act analogous to the scribal role, trumpeting forth the Gospel. The scribe is the author, and in both roles he channels truth, serenely receiving and delivering divine inspiration rather than struggling to reinterpret. Ezra is shown in a similar pose in the Codex Amiatinus, wearing colourful robes and eschewing the desk in front of him in favour of the more intimate process of writing on his lap. 16 The open bookcase behind him claims the scribe as possessor and transmitter of wisdom, performing a similar function to Matthew's angelic trumpet. As I have argued elsewhere, these images are part of an early conception of scribes as a passive conduit, serving the authorial voice by reproducing text. 17 Either as a cause or result of this focus on scribing as a celebrated and meditative conduit, images of contemporary scribes are unusual in the Anglo-Saxon period. The only one of which I know is a small image in the initial B of a Beatus in the Bury Psalter, shown in Figs 6.1 and 6.2. 18 Unlike his more prestigious biblical counterparts, he works over a table and wears a simple, single coloured cowl, with a hood rather than a halo. Tiny, he is contained within the letter as a functional element, rather than occupying a full page like Matthew and Ezra. The green of his robe and blue background draw this small figure into the decorative scheme of the page rather than granting him individuality. This is the scribe as functionary rather than creator or interpreter, perhaps closer to scribes' self-perception as opposed to the creators of the texts they copied.
It may therefore not be surprising that the most attested sensory experience of scribing is the misery it induces. Richard Gameson has collected scribal colophons, the overwhelming majority of which, including all of those which mention sensory experience, record complaints. For instance: He who does not know how to write does not think it is a labour. Writing is done with three fingers but the whole body labours. Pray for me whosoever shall have read this book. It is finished. Through the holy name, your son, our Lord, Jesus Christ.
Clearly, the potential sensory stimulation of an activity is not necessarily enough to render it exciting. I loved my first night shift in a factory: the slightly disjointed feel of sleeplessness and artificial light; smells of hot metal and grease; different languages around me; the feel of rapidly cooling plastic. But my excitement barely lasted the first shift; by the second night, the lack of autonomy took over. Possibly, scribes similarly experienced their work as passive parts of a machine, flawed facsimile-makers with no creative engagement of their own. 20

Difficulty and Drudgery -Perceptions of Scribal Activity
Despite this self-presentation as miserable drudges, recent scholars have been keen to praise scribes, finding them "marvellously creative". 21 In his edition of the Exeter Book, Bernard Muir attributes the arrangement of texts and some adaptation to the scribe, whom he associates with the 'anthologist' who selected texts for inclusion, suggesting that "[i]n another manuscript context these same poems might have received different treatment at the hands of a different anthologist or scribe". 22 The same scribe has been argued to have adapted presentation by Winfried Rudolf, who found that marginal capitals 19  support interpretation of the Riddles. 23 Andy Orchard has found the main scribe of Junius 11 to be fully engaged with Genesis, actively highlighting poetic repetition. 24 Elsewhere, Muir claims Anglo-Saxon texts as inherently "living, dynamic entities that were regularly adapted during the process of transmission". 25 Joyce Tally Lionarons makes a similar argument, but attributes the dynamism to scribing, arguing that "each rendition by an individual scribe represents a unique (re)performance of the work". 26 More cautiously, writing about the Nowell Codex, Josef Klegraf suggests that a scribe could "bring in features of his own 'style' as long as this practice does not interfere with a text's content or 'message'". 27 Considering the same manuscript, focussed on Beowulf, Kevin Kiernan has gone further, ascribing an authorial role to scribes, 28 supported recently by Helen Damico. 29 However, earlier analyses have focussed on the mechanical aspect of scribal work. Scribes are castigated if they seem to have made changes, and admired if they have replicated their exemplars. The ideal scribe for early students of Old English poetry was thus one whose presence could not be detected: production of facsimiles, rather than editions, was the objective for scribes, if not for editors. 30 As Matthew Fisher has argued, this kind of judgement was and still is brought into a moral dimension, with scribes "suspected" of "corrupting" texts. 31 Michael Lapidge discusses this shift, considering the history of emending Beowulf and identifying a movement away from editorial "contempt for the transmitted text". 32 Lapidge supports this more critical approach to scribes' work, seeking to re-establish the distance between scribe and author, and placing the editor firmly in the latter's camp. 33 Douglas Moffat is similarly cynical about scribal capacity to "be like poets". 34 More recently, Lapidge has argued that the scribes of Beowulf failed to understand large parts of their exemplar, questioning their literacy in earlier forms of script. 35 Leonard Neidorf has followed him, suggesting that, with Beowulf, the scribes tried their best but bit off more than they could chew. 36 The Vercelli scribe is likewise found to show "little fluency" and "little intelligence" in copying the prose homilies which dominate the manuscript, and their editor is content to express "contempt" for him. 37 He is repeatedly guilty of "mechanical error" even though he seems to want to be "a slavish copyist at pains to reproduce his exemplar exactly". 38 This all follows Kenneth Sisam's claim that scribes simply cannot be trusted, because they freely altered wording and often did not understand what they were copying. 39 The truth is that scribes vary in their approach to their work, and some may have changed approach when working on different texts. To paraphrase Morrissey, some scribes were better than others; some scribes' projects had more scope for creativity than other scribes' projects. If later medieval practice is any guide, the same scribes may even have had different attitudes to the same text when copying it in different contexts. 40 Scribal inconsistency is mirrored by scholars: as is apparent from this swift and incomplete survey, there is considerable variation in descriptions of scribal activity and what the criteria are for doing it well. And whether it is to be praised or reviled, it is clear that readers and editors of Old English poetry must engage with scribal activity. Such variation in terms used to describe individual scribal performances makes it difficult to analyse scribes, and still harder to access the context in which they worked. 'Mechanical' copying is, for instance, used by Klegraf to mean a scribe who writes in the patterns of his training, unconsciously converting orthography, layout, and script from exemplar forms to those with which he is comfortable. But Katherine O'Brien O'Keeffe, Leonard Neidorf, and perhaps Donald Scragg, use the same term to mean a scribe who reproduces what he sees in an exemplar without intellectually engaging with it or troubling to resolve it into sense. 41 I use it above to mean the process of reproduction without necessarily implying a lack of thought. How should we describe an eighth-century scribe who refreshes some parts of an old text in MS London, British Library, Additional 40165.A.1 to make it more readable? 42 And what is a scribe doing when he intentionally -rather than thoughtlesslyreproduces exemplar forms, such as the scribe who tried to write some phrases in the script of his exemplar in MS Würzburg, M.p.th.f.79, 43 or my findings about the scribes of the Nowell Codex, who varied the forms of initial capitals, apparently echoing those used in their exemplars? 44 What is the difference between this and Scragg's Vercelli scribe, who intended to reproduce his exemplar but utterly failed to do so?
It seems to be the case that a scribe's performance is judged by two criteria. First, what was the attitude to the exemplar(s): exact reproduction or remaking for some contextual reason? Secondly, did the process of the text 40   passing through their hands change that text for better or for worse? The first criterion assesses scribes on their own terms by asking what they thought they were doing in the process of copying. The second judges their results in the context of (intrinsically imperfect) modern understanding of their base text. It should be self-evident that both judgements are based on partial and subjective grounds. It is extremely difficult to assess what scribes thought they were doing with their texts; it is equally difficult to have any secure knowledge of their base texts. But, as I shall briefly suggest below, some such evidence can be gathered and analysed using some rather crude tools.
These two continua of assessing scribal performance can be placed together to produce a matrix, as shown in Fig. 6.3, which assesses both the degree of freedom scribes seem to take, and the extent to which they are improving or damaging their texts. 45 In the middle of the matrix are scribes who precisely recreate what they see: faithful 'transmitters' who create a facsimile copy. A scribe shaping his text creatively can be seen as a 'performer' of that text, akin to the scribal authors that Fisher and others have identified in the later period. A more mechanical 'shaper' works like a modern editor: seeking to present the source text accurately but re-shaped for a new context. A scribe who seeks to copy mechanically, but fails and mangles his text, is 'incompetent'; one who mangles the text while seeking to be creative with it is 'idiotic'. There is no space here to delineate the implications of these different types of activity, nor even to give examples of each case, but it is to be hoped that such a matrix will bring some clarity to discussions of scribal performance. I will, though, briefly explore some creativity that can be seen in codices containing Old English poetry before returning to the initial question of whether any of this activity indicates engagement with the rhythms of texts.

Some Instances of Scribal Creativity
It is clear that we do not yet fully understand what choices scribes made in the presentation of Old English verse. As Francis Newton puts it, "[t]he student of manuscripts must always be open to the possibility of form and artistry". 46 A possible area of scribal artistry is spacing. It has been clear to scholars for some time that word spacing in Old English is not random; it is equally clear that the inconsistency of spacing means that identifying any 'rules' is currently impossible. Thomas Cable noted this difficult situation, admitting that in Beowulf "there is spacing where I cannot explain it, and in other places no spacing where I would expect it to occur", but also noting that some of the use of spacing "forces on us unexpected perceptions of a complex and endlessly teasing aesthetic form". 47   "expressive" choices. 48 A more recent assessment of punctuation follows the same line, finding it "not quite systematic, not quite random". 49 The extent to which pointing should be used to mark separate verses certainly seems to have been an individual preference. In the Exeter Book, at least one later reader was clearly dissatisfied with the sparse pointing provided by the scribes. Marks, usually slashes, indicate verse boundaries through a number  of texts, including Guthlac and Juliana. 50 They are used so regularly that they even occur where the scribe has pointed the verse ending. As shown in Fig.6.4, marks are repeated in the same site, with two lines at different angles between "hondum" and "heofenum"; sometimes they cross one another as in the first line of Fig. 6.5 between "breostum" and "broþor"; sometimes, marks contradict one another in their interpretation of metrical breaks as illustrated by Figure 6.6, where different readers have placed marks either side of "ar". This all suggests that more than one reader participated in this action and that some readers disagreed with one another about where verse boundaries lay. In Junius 11, diverse practice seems to have been read as incompetence rather than dif-50 The marks are discussed by D. MCGOVERN, "Unnoticed punctuation in the Exeter Book", Medium AEvum 52 (1983), pp. 90-99. I am grateful to Jane Roberts for discussing them with me, and to Peter Thomas and the other staff at Exeter Cathedral for allowing me to examine them. Images reproduced by kind permission of the Dean and Chapter of Exeter Cathedral. ference of opinion: a proof-reader corrected Christ and Satan, including the addition of pointing in a number of places. 51 So, while scribes could make decisions about presentation, those choices were expected to fit within an overarching project. Junius 11 required metrical pointing, so failing to include it did not show admirable fidelity, but unthinking reproduction. Scribes are generally identified and dated by the letterforms they use, and yet it is well known that they both could and did vary the forms they deployed. Kiernan suggests that palaeography "is not a test of dating but of chosen style". 52 An obvious instance is in the Exeter Book, where the scribe varies his Square Minuscule hand at key moments. This is particularly noticeable at the end of The Phoenix, lines 667-677, on f. 65v, where the b-verse, or second half of each poetic line, is in Latin. Whenever a appears in these words it is written in the half-uncial oc form, shifting back to the regular square topped minuscule a for Old English words. 53 Similar instances have been found elsewhere. Matthew Hussey showed that the scribe of MS Würzburg, M.p.th.f.79 was working hard to write in uncial script when he was more accustomed to writing in halfuncial. 54 Following Timothy Graham's work on MS Cambridge, Corpus Christ College, 57, 55 Julia Crick has identified a wide number of scribal attempts to archaise their hands, many of which are forgeries. Some, however, represent aesthetic choice rather than deception, demonstrating, as noted in the introduction to this piece, that "some texts needed to be seen as well as heard", and, further, that scribes were sophisticated in responding to their readers' multisensory expectations. 56 Patrick Conner, reviewing such instances, argues that "it is possible for scribes to call forth a hand" as required for a particular project. 57 I have argued elsewhere that small capitals are another scribal feature which scribes used creatively. 58 The first scribe of Beowulf in the Nowell Codex uses minor, non-marginal, capitals quite consistently between two of the prose texts that he works on, differently in the poem Beowulf and differently again from the second scribe of Beowulf. Thus it seems likely that he saw himself as shaping the text of the poem and quite probably the other texts, too. There is, in fact, a quite remarkable degree of clustering of minor capitals in his portion of Beowulf. His average use is of one minor capital every 1.61 sides. It is, therefore, striking that seven occur within four sides at 1.75 per side from folio 163 (BL 166)r to 164 (BL 167)v. 59 These pages contain lines 1491b-1591a, from Beowulf's final words to Hrothgar before he dives into the mere up to the moment when blood from the decapitated monsters flows up through the lake and is seen by the waiting Danes and Geats. That is, the passage with more minor capitals than any other in the text is also the passage which contains the deaths of two monsters. As I have discussed elsewhere, some (but not all) other bursts of use also correspond with monstrous activity. 60 It seems likely that we are seeing a creative scribe responding to and re-presenting his text. It is worth noting in this context that I have not been able to discern any patterns of use in Alexander's Letter to Aristotle, the long prose text that precedes Beowulf, copied by the same scribe and probably coming from the same exemplar. 61 In this text, small capitals do cluster but apparently only around the start and end of the text, rather than across passages with particular meanings. It may, then, be the case, that scribes (or this scribe at least) were more inclined to engage creatively with poetry.
Throughout the four codices that contain the majority of extant Old English poetry, as George Krapp wrote of Junius 11, small capitals "are not systematically used, but when they appear it is almost always possible to see a definite purpose in their use". 62 In Fig. 6.7, I have laid out the frequency of minor capitals in some of the longer texts of the different manuscripts. Where different scribes are thought to have shared texts, I have also shown the frequency of use by the different scribes. I will not use most of the terms of comparison here, but have provided full data sets with figures per manuscript and poetic line as well as per manuscript side in the hope that more work can be done with them in the future. 63 The simple points that I want to make here are, first, that the variation apparent within Beowulf is also apparent in Christ and Satan, the other shared text in this group, although the numbers of sides copied by the three scribes of the latter text vary too widely for this comparison to be secure. Secondly, there are some interesting variations and consistencies across the different manu- scripts. In the first three texts of Junius 11, for instance, the use of minor capitals is broadly consistent, varying only between 1.22 and 1.57 per side, a negligible difference. In the Exeter Book, however, more variation is apparent. While most texts have around three minor capitals per side, the full range is from 2.79 per side in Christ in Judgement (Christ III) up to 7.08 per side in Guthlac A. Quite why the scribe used so many more minor capitals in the Guthlac texts is not revealed by this approach, but is worthy of further study. The general impression is that scribes have 'default' behaviours which vary under certain circumstances. This reinforces the idea of scribes who could be more or less engaged with the process of interpreting and re-presenting their texts. Scribe A in the Nowell Codex was certainly more engaged with a poetic than prosaic text, but more investigation into scribal behaviour across prose and poetry in, for instance, the Vercelli Book, is needed to know if this variation is frequent. Another presentational method that these scribes deploy to varying degrees is coincidence of the end of a manuscript side with the end of a poetic verse, as shown in Fig. 6.8. 64 Most scribes do this on more than half of the pages they copy, and some do it more frequently. This summary is a generalisation: the 'incompetent' scribe of Junius 11's Christ and Satan, and Scribe A of Beowulf do so only about 30% of the time; it is more frequent in the Exeter Book and Junius 11 than in the Nowell Codex and the Vercelli Book. More striking than these low incidences is the very high rate in Junius 11, where the main scribe made this happen on over 80% of pages through Genesis and Exodus. That even the most skilled scribe working hard at his presentation does not coincide verse and page endings on every occasion demonstrates that it was not absolutely necessary, or perhaps that it was challenging to achieve. It is also interesting that the main scribe of Junius 11, when he coincides verse and page ending, overwhelmingly does so for the b-verse, where there is less difference in verse treatment shown by scribes in other manuscripts. That is, in Junius 11 the coincidence of the end of a poetic line with the end of a manuscript page is almost as high as the coincidence of the end of a verse with that of a page. This is in line with the inconsistency in use of the other graphical features discussed above: the argument is not that every scribe was constantly thinking about how to present their text and working to do so in creative ways, but that many scribes sometimes did and that, therefore, these presentational devices had some value for readers.
Along with the general weight of statistical evidence of this phenomenon are a number of incidences where scribes have visibly worked hard to end their manuscript side at the end of a verse. At the end of God's speech to Lot in Genesis, line 2513, the scribe ends page 116, leaving enough space at the end of the line for most of line 2514a. At the end of line 385 of Exodus on page 161 of the manuscript, he does the opposite, writing the last three letters of "stigon" beneath the main ruling rather than splitting it across two pages. Another tactic can be seen at line 1866 of Genesis, page 89, where the last word of the poetic line, "geðreadne", is squeezed into the line, moving the right hand margin much further across than any of the preceding manuscript lines.
A significant weakness in this argument is that it may well be the case that many breaks relate to meaning rather than to metrics and are only "coincidentally metrical in shape". 65 It is certainly true that many pages are designed around meaning: in Junius 11 this is particularly noticeable when gaps are left at the foot of pages when the scribe clearly regarded the 'section' as complete. 66 However, it is also worth noting that in a number of instances, such as line 385 of Exodus, above, the break is not semantic. Peace be with you, true God, and the strength of wisdom, and thanks to you, | seated in glory, for your fresh gifts and for every good thing.
In keeping with the fairly sparse punctuation policy of the Exeter Book, there are few points on the side. I count five before the final manuscript line, the others at manuscript lines 5 (twice), 8, 9, and 14. 68 In keeping with standard practice, a point precedes the start of the speech, which also starts with a minor capital. What is interesting here is that the manuscript line ends with line 623a and, where there would be space to write "þrym", the first part of the first word of 623b, the scribe instead places a point. There is no break of sense here: there can be little doubt that both the point and the break between pages relates to the coincidence of verse and side ending.
It is also worth noting that if scribes sought to bring sections of meaning to an end at the close of pages, they were not very good at it in prose. As noted 67 The Phoenix, lines 622-624. Bars inserted in Old English and translation are mine, showing page breaks. 68 Points follow poetic lines 596a and b, 600b, 602b, 610b. The last two are also punctuated by Krapp and Dobbie. above, the first scribe of Beowulf also copied three prose texts. Excluding the final page of each text, there are strikingly few instances where a unit of meaning (a clause, or sentence, or section) comes to an end at the end of a page. It never happens in the fragment of St Christopher. There are three instances in Wonders of the East, one on the first page, possibly indicating that it was a desideratum. 69 There is not a single instance in fifty sides of Alexander's Letter to Aristotle. 70 Wonders is a highly structured text, divided into short sections mostly accompanied by images, which may have made it easier or more desirable to connect semantic breaks with page endings. However that may be, it remains the case that in seventy-three sides of prose, the scribe coincides the ending of just three with a break in a unit of meaning such as a sentence or a section, at a rate of 4.11%. 71 While the proportion of the end of poetic sides coinciding with a metrical or semantic break is hardly compelling in Scribe A's work, it is overwhelming in comparison with the rate of such breaks in his prose. Taken together with the evident intention to produce such breaks in Junius 11, it is clear that scribes sometimes organised their copying around the content of texts and that they seem to have found it easier to do so when those texts were structured metrically, and, further, that they sometimes (as in the example from Exodus, above) organised their copying on purely metrical criteria. Another way of putting this, in the context of the foregoing discussion, is that scribes were more inclined to creatively reshape their texts if they were poems.

Did Scribes "Whistle While They Worked"?
It is certain that some scribes thought about the process of copying, and that they found significance in their own role and the process of textual reproduction. It is also certain that at least some scribes actively understood the copying of Old English texts as the making of texts for aural reception. Scribing could be, and often was, a multi-sensory experience, engaged with the tools 69 On ff. 95(97) (BL 98)v, 100(96) (BL 103)r, and 103 (BL 106)r. 70 The modern editions make it look as though there is a sentence ending at the foot of f. 113(121) (BL 124)r, with a new sentence starting 'Eall ...' on the next side. It is not marked by the scribe, who points before the editorial break, and the last letter of the first word on f. 113(121) (BL 124)v is not l. 71 There are seventy-six prose sides, but three are the final sides of pieces and I do not include them here. of the trade and with the process of book-making. And it is certain that, while some scribes conceived of their role and performed (more or less effectively) as mechanical reproducers of text, many did not and instead engaged creatively with their texts. But, to return to the questions asked at the start of this paper, is there evidence of scribes engaging with the metrical rhythms of Old English poetic texts? Is it possible to imagine a scriptorium where scribes whistled as they worked, hearing the words and verses they wrote? The short answer is that the evidence is inconclusive. Evidence of such individual and transitory experiences is sparse; proof is virtually impossible. But as work continues to consider the evidence for how scribes worked and what they expected readers to do with their texts, we draw closer to understanding some of the multiple contexts of scribal activity, and hence to appreciating the multiple sensory experiences provided by texts in Anglo-Saxon England.