In the nineteenth century, John Romilly Allen confidently claimed that the iconography of the Crucifixion with the robed or ‘fully draped’ Christ was a phenomenon of Celtic art, found in Scotland, Ireland and Wales, distinguishable from the ‘Saxon’ type in which Christ wore a loin-cloth. Other features of the Saxon type were the presence of the sun and moon above the arms of the cross, instead of angels as in Ireland; and the figures of the Virgin and St John at the foot of the cross, without the spear- and sponge-bearers, the latter pair appearing only exceptionally at Alnmouth, Northumberland; Aycliffe, County Durham; and Bradbourne, Derbyshire. Clearly two different versions were identified in this analysis, but no attempt was made to clarify the chronological relationship between the examples cited, and only the geographical distribution of a small number of examples was considered. Romilly Allen's confidence in distinguishing ‘Celt’ from ‘Saxon’ on the basis of art styles, even for the pre-Viking period, is not always shared today, as the continuing discussion of the origins of several important manuscripts shows. The terms ‘Insular’ and ‘Hiberno-Saxon’ used to describe much of the art from the sixth century to the eighth underline die perceived difficulties.

My article begins in true medieval fashion with a retraction. A few years ago I argued that The Wanderer and The Seafarer involved two imaginary speakers rather than one. Ever since the publication of this argument I have been assailed by doubts of its validity, not only because most persons whose judgement I trust in such a matter have been either politely noncommittal or vociferously hostile, but because I myself, in returning again and again to the poems, have been less and less convinced that their texts would support my reading. The decisive moment came, I think, when I read P. L. Henry's strong reinforcement, from the Celtic side, of Dorothy Whitelock's view of The Seafarer as the monologue of a religious ascetic who had chosen exile on and beyond the sea for the love of God – a peregrinus pro amore Dei of a sort well known in the British Isles from before Bede's time to Alfred's wherever Celtic Christianity had taken root. In thus returning to Professor Whitelock's interpretation I found myself completely convinced of its superiority. Since that time I have altogether abandoned my former view of The Seafarer, and with it my view of The Wanderer also, and I welcome the opportunity to say so.

Oral literature research entered Old English literature through Albert Lord's 1949 dissertation, eleven years later to become The Singer of Tales, and Francis P. Magoun, Jr's ensuing essay, ‘The Oral-Formulaic Character of Anglo-Saxon Narrative Poetry’.1 As I have pointed out elsewhere,2 these two scholars were not the first to identify and discuss the recurrent phrase or ‘formula’ in the poetry; German Higher Criticism of the nineteenth century had analysed the use of commonplaces of diction (or Parallelstellen) to try to determine authorship and to establish the text of various poems. What Lord and Magoun originated was the idea of an explicit and necessary connection between the formula, defined by Milman Parry as ‘a group of words regularly used under the same metrical conditions to express a given essential idea’,3 and a poem's orality. Following the lead of Parry's ground-breaking analyses of Homeric epic and the Parry–Lord field work on South Slavic oral epic, they and scholars following them reasoned that the source of formulaic structure lay in a tradition of oral verse-making and that formulaic phraseology was a kind of poetic idiom fashioned over generations by bards responding to the continual pressure of composition in performance. Only if his mind were well stocked with phrases of metrical shape – if, in short, he had learned his poetic language well – could an oral poet fluently tell his tale in the form of traditional verse.

Frankfurt, Stadt- und Universitätsbibliothek, Fragm. lat. I. 56, a single detached leaf written in Anglo-Caroline minuscule of mid-eleventh-century date and containing a number of neumed metra of Boethius, was recently brought to the attention of Dr M. T. Gibson1 by the Keeper of Manuscripts in Frankfurt, Dr Gerhard Powitz.2 Dr C. Page identified it as part of the ‘Cambridge Songs’ manuscript, Cambridge, University Library, Gg. 5. 35 (fols. 432–41). The leaf throws valuable new light not only on the subject of musical notation in late Anglo-Saxon England, but also on the composition of one of the best-known collections of medieval Latin lyrics.3

Having just killed the dragon, and with death ‘ungemete neah’ (2728b; ‘exceedingly near’), Beowulf begins to speak, remarking first on his lack of a son:

Nu ic suna minum syllan wolde

guðgewædu, þær me gifeðe swa

ænig yrfeweard æfter wurde

lice gelenge. (2729–2732a)

The language here seems almost neutral, but at least two of the terms are suggestive. The ‘guðgewædu’ (‘war-garments’) that he wishes to leave to his son recall most notably the elaborate gifts of weapons and horses – including the saddle that the king used in battle (1037b–1043) – that Hrothgar had given him following the fight with Grendel, implicating him, as John M. Hill and Stephanie Hollis have argued, in the Danish succession. Similarly, in using ‘yrfeweard’ (‘guardian of an inheritance’), he calls attention to the king's role in protecting the tribe's wealth, although the poet may expect the audience to hear more in the compound since ‘weard’ appears often in the final part of the poem in descriptions of the dragon. Thus, even though Beowulf does not mention this failing again, his remark draws together a theme not only signficant in the Geatish section of the poem, but also present from the opening descriptions of Scyld's reign: part of a king's responsibility is to see that the succession is secure in a son.

The Ordinance concerning the Dunsæte sets out regulations for dealings between the English and Welsh in some part of the frontier between these two peoples. The text is widely assumed to be from the second quarter of the tenth century, but this article argues for a late-tenth- or eleventh-century date. The Ordinance envisages trade and prescribes procedures to settle disputes: it thus reveals cordial contacts between those dwelling along the frontier. This offers an alternative perspective to the focus on kings and conflicts found in many modern accounts of relations between the English and Welsh in the early medieval period.

The classification of The Phoenix as a Christian allegory has obscured the perception that the poem may contain a single symbolic vision, for the attention given the poem has always centred on the apparent allegorizations drawn from the Lactantian De Ave Phoenice. The long description of paradise at the beginning, the phoenix's journey from paradise, its fiery death and resurrection – elements derived from Lactantius and expanded by a clearly Christian reading – invite the exegete to untangle a carefully woven allegorical web. The exegetical approaches are, in some ways, justified. Even Lactantius's poem, which at no time refers to a system of Christian theology, is legitimately subject to this kind of inquiry, for the phoenix, standing by itself, had been a symbol charged with Christian meaning since the earliest patristic writers. Augustine's words may be taken as a summary statement for the Christian interpretation: ‘Quod enim de phoenice loqueris … Resurrectionem quippe ilia significat corporum.’ In practice, however, strict allegorical readings of The Phoenix have not proved very helpful. An ever-changing perspective in the poem makes it difficult to discover a logically coherent and consistent pattern of allegorical meaning, and attempts to find such a pattern have led either to disappointment with the poem or to an exegetical system so rigid that it falsifies the poem itself.

The literature on watermills is extensive, and includes many papers on what is usually called the ‘horizontal’ mill, that is to say mills with a horizontal wheel and a vertical shaft, as against those with a vertical wheel and a horizontal shaft. It is here proposed that the type under discussion in this paper, sometimes called the ‘Norse’ or ‘Greek’ mill, should be called the ‘horizontal-wheeled’ mill, and the other the ‘vertical-wheeled’ mill.

In the introduction to his edition of the New Minster Missal (Le Havre, Bibliothèque municipale, 330), Derek Turner tells us that the manuscript he is editing ‘is one of the five fairly complete mass-books of probable English provenance remaining from before 1100, and the oldest true missal’. By ‘missal’ he means, of course, that Le Havre 330 is a mass book of the sort which usually contains a suitably integrated ordering of texts and chants of the sacramentary, the lectionary and the gradual according to the various traditional propers and ordinaries of the temporal and sanctoral arranged to accompany the church year; normally there are also some votive masses after the main corpus of the book.

Recent research has focused attention on the ‘Lotharingian connection’, that is to say the close links between the English and German churches in the middle of the eleventh century. Its best known manifestation is the presence of a significant number of German (mainly Lotharingian) clerics at the English royal court. This phenomenon seems to have its origin in the reign of Cnut (1016–35), and the purpose of this paper is to shed light on two aspects of Cnut's contacts with Lotharingia. First, an explanation is advanced for Cnut's baptismal name, Lambert. Secondly, attention is drawn to a hagiographical source which provides evidence for a visit by Cnut to Cologne.

The nineteen-line Old English poem known as Wulf and Eadwacer has proved a notorious lodestone and analytical trap for critics; and still another interpretation of it may seem futile, if not presumptuous. Nevertheless, I believe no more firmly in my interpretation than others have believed in theirs: that is, I am no less confident that it will clear up most of the verbal and situational mysteries the poem presents; that it will enable us to see the lyric as structurally whole; that it will help us appreciate even more its aesthetic qualities; and that it will gain a critical consensus. To achieve such modest goals, I shall have to consider assumptions about the mind-set of the Anglo-Saxon audience as well as the poem's structure, diction, tone and imagery. It will thus be well to have the poem before us, and I venture a poetic translation which I shall comment on in due course.

According to the treatise which he presented as a New Year's Gift to King Henry VIII in January 1546, the English antiquary John Leland (c. 1503-52) received ‘a moste gracyouse commyssion’ from the monarch in 1533 ‘to peruse and dylygentlye to searche all the lybraryes of Monasteryes and collegies of thys your noble realme’. As he travelled from monastery to monastery he compiled lists of books, some brief, some considerably more thorough. Leland had a strong interest in pre-Conquest writings and his lists testify both to lost exemplars of known texts and to otherwise unattested writings from the early period. His principles of selection were not always consistent, however, and on occasion he omitted manuscripts of historical or literary interest which he must have seen: in no sense was he setting out to be a cataloguer as such. Ultimately his enterprise was a bibliographical one and he envisaged a volume in four books, De uiris illustribus siue de scriptoribus Britannicis, of which ‘the seconde is from the tyme of Augustyne, unto the aduente of the Normanes’. The original draft of the De uiris illustribus was composed around 1535/6 – it was in part a response to Polydore Vergil's Anglica Historia published in Basel by John Bebel in 1534 – but after this burst of activity there was a gap of almost ten years before Leland's next major phase of composition.

Literary history emerges when critical readers in sufficient number move beyond primary recognition of individual texts into a secondary awareness of a scheme, a sense of the connections that exist between these texts.1 Literary history considers the development of a whole body of literature, tracing multifarious influences and innovations through time. In the course of Anglo-Saxon studies the slow and sporadic reappearance of the literary remains resulted in the late nourishing of a schematic or historical overview. As Wellek reminds us, ‘the antiquarian study of Anglo-Saxon remained…outside the main tendency towards literary history’2 that occurred in late-seventeenth- and early-eighteenth-century England. So, too, the special quality of Old English poetry itself contributed to the laggard creation of a history. It is difficult to map the path of a literature in which all dating is only good guessing and in which a tenaciously conservative oral—formulaic style makes attempts at suggesting influence hazardous.

The Old English poem Andreas deserves a reading entirely on its own terms. For years we have been studying it in various distracting ways: as if it were a deliberate and feeble imitation of Beowulf (which it is not);1 or trying to attach it to the Cynewulfian canon;2 or taking it as a compendium of patristic lore (which hardly any Old English poems are);3 or we may see it as an example of an inherently inferior medieval genre, hagiographical romance; or (most often) as an embarrassing misapplication of the heroic style to the wrong subject. Clearly the poem invites us to make these classifications, but its own noteworthy achievements have been generally overlooked. We need a straightforward frontal attack by the critic: just what is this poem up to, line by line? The best guide in the process seems to me to be the Latin source, to which we have a fairly close approximation in the Recensio Casanatensis.4 Careful consultation of this source lets us see what forms the poet's imagination imposes on his given material and gives us the clearest idea of the kind of poet he is.

A group of half-lines in Cynewulf's poetry that take the form þurh + demonstrative pronoun + adjective + gesceap/gesceaft constitutes a distinctively Cynewulfian realization of a more widespread formulaic system (x X gesceaft). Only Cynewulf substitutes (in two closely similar contexts) gesceap for gesceaft, and Cynewulf consistently uses both words to refer to a specific ‘created thing’ instead of to ‘creation’. Within the system as a whole, the specific referent of gesceaft in the sense ‘creation’ (heaven, or earth, or all of creation) is often clarified by deixis, specifically in the type of demonstrative pronoun that modifies the word. Cynewulf's choice of the distal pronoun þæt likewise indicates that the apparently ambiguous referents of gesceap/gesceaft in Juliana 273b and 728a and Elene 789a are heavenly or unearthly things, and analysis of context as well as Christian-Latin analogues supports a specific identification of the referent in each case. While availing himself of an existing formulaic system, then, Cynewulf was innovative in its metrical realization as well as in its semantic application.

Readers of Beowulf are now in apparent agreement that the frequent Christian references in the poem fit easily and naturally into their contexts and give every sign of being simply an unconscious part of the normal language and thought of the poet. We must make the customary proviso, to be sure, that by ‘poet’ we have in mind the final reteller of some older pre-Christian story or stories now beyond recovery. The old Monkish Interpolator who used to be hauled out and execrated by scholars for smearing his despicable Christian Colouring like graffiti over a noble pagan monument has vanished like the ghost he always was. He has, however, now been replaced by equally busy Monkish Extrapolators working tirelessly to insert large sections of the Patrologia Latina somewhere between the lines of Beowulf. In doing so, they take advantage of the undeniable fact that the early Middle Ages provided an ample supply of patristic material, to argue (I think indefensibly) that it must have been put to use – that no vernacular poet could have fought off the urge to wrap any plain tale he had to tell in rich folds of typology, allegory and deeper meanings.

Oxford, Bodleian Library, Junius 27 (S.C. 5139), the Junius Psalter, was written, Latin text and Old English gloss, probably at Winchester and presumably during the reign of King Edward the Elder. Junius 27 is one of the twenty-nine complete or almost complete psalters written or owned in Anglo-Saxon England which have survived. (In addition to these twenty-nine complete psalters, eight minor fragments of further psalters are still extant.) This substantial number of surviving manuscripts and fragments is explained by the paramount importance of the psalms in the liturgy of the Christian church, both in mass and especially in Office. Junius 27 is also one of the ten psalters from Anglo-Saxon England bearing an interlinear Old English gloss to the entire psalter. (In addition there are two psalters with a substantial amount of glossing in Old English, though not full interlinear versions.) Since our concern in the first part of this article will be with the nature of the Old English glossing in the Junius Psalter, and its relationship to other glossed psalters, it is appropriate at the outset to provide a list of the psalters in question. At the beginning of each of the following items I give the siglum and the name by which the individual psalters are traditionally referred to by psalter scholars. An asterisk indicates that the Latin text is a Psalterium Romanum (the version in almost universal use in England before the Benedictine reform); unmarked manuscripts contain the Psalterium Gallicanum. For full descriptions of the manuscripts, see N. R. Ker, Catalogue of Manuscripts Containing Anglo-Saxon.

Chapter 73 of the laws ‘þe Æðelbirht cyning asette on Agustinus dæge’ reads ‘Gif friwif locbore leswæs hwæt gedeþ, xxx sell' gebete.’ The sole manuscript evidence for the laws of Æthelberht is in the early-twelfth-century Textus Roffensis; since this is so late, the manuscript word division is not necessarily significant. For what it is worth, two of the three crucial words are written locbore and les wæs, but it is not absolutely clear whether friwif is intended as one word or two. The elements are divided by the smallest of spaces. Friwif and locbore do not occur elsewhere in the laws; les wæs occurs also at Æthelberht, ch. 3, written lyswæs. Comparatively recent texts and translations can be found in editions of the Anglo-Saxon laws by Thorpe, Liebermann and Attenborough, and, translation only, in Dorothy Whitelock's English Historical Documents (probably the most widely used and certainly the most readily accessible of the four). The earliest of these editors, Thorpe, avoids the difficulties by leaving certain problematical words in the original: ‘If a free-woman loc-bore commit any leswe let her make a bot of xxx shillings.’ Liebermann offers: ‘Wenn eine Freie, eine Lockenträgerin, etwas Unzüchtiges thut, büsse [ihr Schander ihrem Vormunde] 30 Schll’, where, as Attenborough subsequently and charmingly explains, ‘Liebermann understands as the subject of gebete not the woman, but the man with whom she misconducts herself.’ Attenborough's own translation agrees grammatically with Thorpe's: ‘If a freeborn woman, with long hair, misconducts herself, she shall pay 30 shillings as compensation.’ Whitelock has a minor variant: ‘If a freewoman, with long hair, commits any misconduct, she is to pay 30 shillings compensation.’

In the process of literary interpretation no critic can afford to overlook the rôle of the poetic narrator. While poetic and narrative statements (as it is commonly argued) designate their meaning largely in accordance with the conventions of language and literary discourse, linguistic criteria alone cannot determine the attitude of the speaker towards what he says; and this attitude constitutes a crucial element in the meaning of the statement as a speech act or utterance. Indeed, as users of language, all of us habitually include considerations of speaker intentions in our standard operations of interpreting as well as producing discourse. Can the speaker be trusted? Does he speak ironically or sincerely? Is he trying to achieve some aim in relation to the hearer other than that which his act of communication ostensibly purports? Entailed in any act of communication, this dimension of interplay between speaker and statement is inevitably involved in literary discourse as well, since obviously we do not always take literary statements at face value. One of the primary tasks confronting the literary critic, then, in Old English poetry or any other body of work, lies in determining the character of the narrator and the parameters of his functioning.

As in other societies, adultery was a punishable offence among the Germanic peoples. Although it is a topic which has commanded considerable attention, it has been given attention not so much because it deals with family law and its significance to social history, as because it concerns the treatment of women. But closely related to the question of women, of course, is that of how men view each other. Even as early as Tacitus, evidence exists that Germanic women were treated with respect, and were subject to the protection or mundium of male relatives. Although exaggerated, the account in the Germania gives us some understanding of the role of Germanic women in respect of betrothal, marriage and family life. But it also leaves us with questions to which we most likely will never find answers.