The Nowell Codex and Its Texts

As will become apparent, what I will usually call the ‘Nowell Codex’ is not a straightforward artefact: even its title is contested. I am following Kemp Malone, who first used that name.¹ The shelfmark is London, British Library, Cotton ms Vitellius A. xv (Second Part). It is more widely recognised as ‘the Beowulf manuscript’.² The difficulty in using the shelfmark is partly its cumbersome nature, and partly that what I call the Nowell Codex is the second half of that volume, having been bound together with the Southwick Codex.³ Using one name makes confusion with Cotton Vitellius A. xv as a whole inevitable; I will use ‘Vitellius A. xv’ only when referring to the composite book. The two parts were probably united between 1628 and 1638 by Richard James, Robert Cotton’s librarian.⁴ His motivation remains unclear since the Southwick Codex, a twelfth-century collection of Old English religious texts, seems to have little direct connection with the Nowell Codex.⁵ Although interesting in its own right, the Southwick Codex has been little studied.⁶

For obvious reasons of economy and the significance of the text, ‘the Beowulf manuscript’ is the term most commonly used by non-specialists and indeed in the main title to this volume. The name is useful because it calls attention to the most famous text in the manuscript. But this in itself makes its value questionable: it is not clear that the compiler(s) of the Nowell Codex saw Beowulf as its focal piece and, as I will argue in this study, there is much to be gained from reading the poem in the textual context it is given here. To make matters more confusing, ‘the Beowulf Codex’ is used by Kiernan as a term specifically for the pages containing that poem: it is (or was at the time of his original publication) a significant element of his argument that Beowulf formed a codex by itself.⁷

Unfortunately, similar arguments can be used against ‘the Nowell Codex’ as a title. Kiernan calls it a “confusing new name”, pointing out that if few scholars know what Vitellius A. xv is, fewer still have any idea about the Nowell Codex.⁸ It is so called because Laurence Nowell’s name appears on what is now its first page, along with the date 1563.⁹ This, however, is about two-thirds through the first text – a narrative of Saint Christopher’s passion – so the opening material at least must have been lost before Nowell signed what was presumably a new possession. The original volume was, in fact, configured differently at start and end as I will discuss in Chapter 2. In short, although ‘the Nowell Codex’ describes the material I am considering, the name does not represent the manuscript as it was conceived or produced. This ‘original’ intention is the focus of this book, but accessing it is challenging, for almost nothing is clear-cut about this eleventh-century compilation.

As it now stands, the Nowell Codex contains five texts, two of which are incomplete. These are not titled by the scribes, and, like the manuscript, have sometimes been given different titles by scholars. I will call them

The Passion of St Christopher [henceforth St Christopher];

The Wonders of the East [henceforth Wonders];

The Letter of Alexander to Aristotle [henceforth Alexander];

Beowulf;

Judith.

Foliation and Gatherings

Vitellius A. xv came into the possession of the British Museum in 1753, when Robert Cotton donated his manuscript collection, forming the basis of what would become the British Library. While permanent storage was being sought, the collection was stored at Ashburnham House. In 1731, the building burned to the ground, destroying some manuscripts and damaging many others.¹⁰ Though Vitellius A. xv was less damaged by fire and water than some of the collection, most pages are darkened, all bindings and threads are lost, margins are burnt away, and many words and letters have been lost. Creating a major challenge for studies such as this, the fire also destroyed the stitching of the pages and all other straightforward evidence about how the book was put together. The manuscript survived into the digital age thanks to careful protective work. Each page was placed into a stiff paper frame and these were rebound. Unfortunately, the frames cover up many letters and parts of letters, some of which have since been recovered through the use of digital photography with ultra-violet light. At different points in the rebinding process, most likely shortly after the fire, two gatherings were swapped over and some individual pages moved out of position. In addition, as with many medieval manuscripts, Vitellius A. xv has been subject to various reorganisations, with additional pages dividing the two component codices inserted and removed at different times.

As a result of all of this activity, and the errors that often arise when numbering pages, the pagination of the Nowell Codex is even more challenging than its name. Kiernan has identified at least six distinct foliations:¹¹

1. i.c.1630, under the Cotton librarian, Richard James;
2. ii.1703, by Matthew Hutton, during the work of Humphrey Wanley’s corrective committee;
3. iii.1793–1801, probably by Joseph Planta – this is the foliation Kiernan follows, and that written on the folios themselves;
4. iv.before 1845, pencil numbers in the upper right corners of the paper frames, before Henry Gough rebound the manuscript in that year;
5. v.1845–1884, on the lower right hand corners of recto frames, by an archivist, which includes blank paper leaves inserted by Gough and which was perhaps intended to supplement that of 1845;
6. vi.1884, the final attempt, intended to clarify confusion between the fourth and fifth foliations: this is followed by the British Library.

This sequence of re-numberings means that each recto bears a baffling set of numerals. Kiernan’s proposed foliation system is cumbersome for some pages, but makes good sense, because it identifies the ‘manuscript foliation’ number (Planta’s work from between 1793 and 1801), and gives the British Library’s ‘official’ number (which uses the 1884 foliation) afterwards in brackets; the latter remains important as it is used by many scholars and by the most recent digital facsimile.¹² More significant difficulties come with the third and fourth gatherings, which were swapped around at some point in the manuscript’s history of rebindings. Kiernan’s system puts the current location of these pages in brackets, with where they ‘should’ be as their main folio number. So ‘folio 107(115) (BL118)’ is currently the 115th leaf in Vitellius A. xv. If the gatherings were reordered according to their content, it would be 107th. In the British Library system it is 118th. To add to this confusion, two folios were misplaced when Planta foliated the manuscript. This means that Kiernan’s general rule of using the number written on the pages cannot be followed: under his system these anomalies are folio 147A(131) (BL149) and 189A(197) (BL192) respectively, while there is no folio 131 or 197.¹³ Using Kiernan’s system often makes discussion of the manuscript look more complex than it is, and this will be particularly noticeable in my central chapters (2, 3, 4, and 5). But I have chosen it because it ensures that, whichever pagination readers are using, cross-reference is as straightforward as possible. Using this system, my observations and analyses can be compared more readily with other discussions and with the manuscript itself.

Other Terms and Names

To aid clarity, it is appropriate to note here my use of some terminology. Because discussions of the Nowell Codex are often complex, I seek to use most terms with more precision and consistency than is usually demanded of them. ‘Folio’ is used interchangeably with ‘leaf’ to mean a single page, both recto and verso. ‘Side’ is not synonymous with ‘leaf’ and means either a recto or verso. ‘Sheet’ is used to mean a bifolium: the folded piece of parchment that formed two leaves, four sides. In the Nowell Codex, as all leaves have been slit and set separately in frames, it is not always certain which folios were originally part of the same sheet. ‘Gathering’ is generally preferred to ‘quire’. Where I use the latter term it is in its precise sense of a group of four bifolia sewn together to produce an eight folio, sixteen page set. ‘Gathering’ means any group of bifolia sewn together: the gatherings in the Nowell Codex range from three sheets (six folios; twelve pages) up to five sheets (ten folios; twenty pages) in size. The amount of poetic text written on a side is given in poetic half-lines.¹⁴ This follows the standard approach to measuring the density of scribal work, but uses half instead of whole poetic lines as a finer measure of quantity.

The codex was written by two scribes. A great deal of the discussion in this study is an attempt to engage with the interests and interpretations of these scribes, whom I will take to be men for the sake of convenience. No evidence has been found of either hand in other books or documents. Nor is there any clarity about where or exactly when they worked. One copied the prose texts St Christopher, Wonders, Alexander and the first two-thirds of Beowulf. The other copied the remainder of Beowulf and Judith. The scribes have been variously identified as (1) and (2),¹⁵ S1 and S2,¹⁶ the A-Scribe and the B-Scribe,¹⁷601 and 602,¹⁸ and Scribe A and Scribe B.¹⁹ I follow this last convention because of its greater recent currency.

Wonders often does not give specific names to its creatures and peoples, or uses names now not in common usage. Thus, for instance, no names are given for the people with heads in their chests, nor for the long-eared tribe, respectively known, from Pliny, as Blemmyae and Panotii in parallel texts and discussions. Similarly, the half-man half-horses well-known today as Centaurs are named as homodubii (“doubtful people”), and the dog-headed people more often known as Cynocephali are called conopenae. In keeping with the current general practice of scholarship, I will use the classical name where it exists, and give the name used in Nowell where it varies from this in the first reference. For consistency, all of these tribal or species names are capitalised.

Apart from in Chapter 1, where they will be from the relevant standard edition, quotations from the Nowell Codex texts will be transcribed from the manuscript unless otherwise stated.²⁰ In addition to the use of letters commonly used in editions of Old English texts (æ / Æ, þ / Þ, ð / Ð), this includes the use of a crossed thorn (ꝥ) where the scribes use it to indicate þæt; vowels with abbreviation marks (most frequently ū), usually to indicate following m; the occasional other letter with a bar when the scribes use it to indicate a following letter, such as ḡ for -ge-; wynn (ƿ) for modern w; the Tironian sign ⁊ which both scribes use as an abbreviation for and. When relevant to the discussion, I will format quotations as they appear in the manuscript.

Finally, Anglo-Saxon and Norman names will be given in their most commonly used modern form, with a brief discussion of naming conventions when they first occur if relevant. For Scandinavian names, I join Timothy Bolton in following the convention of using Old Norse (Old Icelandic) forms for people whose main sphere of action was in Scandinavia and Anglicised forms for those who operated primarily in English or international spheres.²¹ Thus, for instance, Knútr (inn ríki) is always Cnut (“the Great”), whereas his father is Sveinn Haraldsson.