Introduction

The short medieval lyric known by its opening words as Hebban olla vogala takes up an outstanding place in the history of Dutch literature.Footnote 1 Frits van Oostrom (2006, p. 107) notably hails it as “het beginakkoord van onze literatuurgeschiedenis” (‘the opening chord of our [Dutch] literary history’). But not only is there still no definitive reading of the text—there has even been some uncertainty as to whether Hebban olla vogala is, in fact, Dutch. It only survives as part of a badly faded pen trial from the late eleventh century, copied into an Old English homiliary then kept at the cathedral priory of Rochester in Kent (today Oxford, Bodleian Library, MS Bodley 340). While this pen trial was long considered West Flemish after its discovery in 1932 (Cotman & Taeldeman, 2003; Schönfeld, 1933), the question of language was reopened when Luc De Grauwe (2004) suggested that its linguistic forms could also match the Kentish dialect of Old English. Since then, a new consensus has emerged, which holds that Hebban olla vogala is a linguistic blend and may even reflect a Dutch-speaker’s attempt to write in English (De Grauwe, 2008; Dekeyser, 2007; Louwen, 2009). Yet although these recent reassessments have done much to nuance the text’s linguistic classification, they do not make direct use of the manuscript. Thanks to the digital imaging techniques available to the Bodleian Library, however, it is now possible to reconstruct the text of Hebban olla vogala with some degree of certainty. This reveals a poem that seems to make deliberate use of the similarity between dialects of Old Dutch and Old English—a poem that is not only part of Dutch literary history, but that should also, and for the first time,Footnote 2 be considered in the context of English literary history in the aftermath of the Norman Conquest.

Since Erik Kwakkel (2005; 2013, pp. 242–243) established that the scribe of Hebban olla vogala was probably trained in the Low Countries, it has been clear that the pen trial must be seen in the context of cross-Channel contacts (Van Houts, 2006a; Van Oostrom, 2006, p. 98). But the mutual intelligibility of late Old English and Old Dutch, which the poem appears to exploit, highlights how easily such literary contacts might have been overlooked due to the similarity of the two languages. This similarity also challenges the often tacitly assumed status of English as a purely local language in post-Conquest England, and it points to an additional current of influence, besides Anglo-Norman, on early Middle English literature. None of this, of course, takes away from the text’s importance to the Dutch literary canon. Rather, in contributing to a growing interest in Anglo-Dutch relations in the medieval and early modern period (see Levelt et al., 2023, pp. 2–4; Levelt & Putter, 2021), it additionally encourages us to look across national boundaries at transnational, multilingual literary exchanges. In this way, Hebban olla vogala finds its place not only in Dutch but also in English literary history, as a testament to their interaction in the late eleventh century.

The Text and Its Form

MS Bodley 340 is the first volume of a two-volume collection of Old English homilies, whose second volume survives as Oxford, Bodleian Library, MS Bodley 342. Probably produced at St Augustine’s Abbey, Canterbury in the early eleventh century, the homiliary was later kept at St Andrew’s Cathedral Priory, Rochester from the mid-eleventh century.Footnote 3 In the last quarter of the eleventh century, Hebban olla vogala was then copied out on the back of MS Bodley 340’s last folio (169v), among other Latin pen trials and along with a Latin translation (Fig. 1).Footnote 4 The script’s angular Continental aspect clearly distinguishes it from more rounded English hands, and Kwakkel (2013, pp. 242–244) notes its similarity to an inscription recording the donation of an eleventh-century manuscript to St Peter’s Church in Utrecht (Leiden, Universiteitsbibliotheek, VLQ 74, fol. 71r), leading him to conclude that the scribe must originally have been trained in the Low Countries. It is assumed this scribe came to Rochester via Normandy in or after 1083, when Bishop Gundulf of Rochester replaced the cathedral priory’s five English canons with more than sixty monks from Norman abbeys (Kwakkel, 2005, pp. 10–14; 2013, pp. 234–235).

Fig. 1
figure 1

Source: © ARCHiOx/Bodleian Libraries, University of Oxford: albedo created by John Barrett with the Selene Photometric Stereo System

MS Bodley 340, fol. 169v (detail).

Full size image

The exact reading of the vernacular pen trial, however, has long been uncertain, and various interpretations have been offered over the last century (see Louwen, 2009, pp. 64–65). Kenneth Sisam (1933, p. 11), who discovered the inscription in 1932, only deciphered the words “hebban olla vogala nestas hagunnan hinase hi / anda thu” (‘all the birds have begun nests, except for me and you’). An ultraviolet photograph published by W. H. J. Caron (1963) allowed for a much improved reading of “Hebban olla uogala nestas hagunnan hinase hic enda thu uuat unbidan uue nu” (‘all the birds have begun nests, except for me and you – what are we waiting for now’).Footnote 5 But the issue was confused with the inclusion of the text in Maurits Gysseling’s Corpus van Middelnederlandse teksten (1980, p. 130), in which it appears in a different form, as “Hebban olla uogala nestas hagunnan hinase hi[c] / [e]nda thu uu[at] unbida[t] g[h]e nu” (‘all the birds have begun nests, except for me and you—what are you waiting for now’). This reading, which introduces a divergence in meaning between the vernacular pen trial and its Latin translation, is notably followed by De Grauwe (2004), but it is not borne out by the manuscript evidence. Most recently, Kenny Louwen (2009, p. 77) reconstructed the text more correctly as “Hebban olla uogala nestas hagunnan hinase hic enda thu uuat umbidan uue nu” (‘all the birds have begun nests, except for me and you—what are we waiting for now’). Yet this is a largely linguistic reconstruction, exclusively based on facsimiles (Louwen, 2009, p. 65), which can only speculate as to the exact forms of anda or enda (‘and’), or of unbidan or umbidan (‘wait’).

A more secure reading of the pen trial in MS Bodley 340 is now possible thanks to the Bodleian Library’s digital imaging facilities. Multispectral imaging, the photography of the manuscript under different wavelengths of light, produced the best results.Footnote 6 The pen trial was copied out in iron-gall ink,Footnote 7 which is badly faded but can most effectively be recovered under ultraviolet light (see Mrusek et al., 1995, p. 74). The clearest picture was accordingly taken under ultraviolet light, with a 500 nm filter and the aperture set to F8 (Fig. 2), which even yielded some improvement on the earlier ultraviolet photograph published by Caron (1963). This is the best image currently achievable with the equipment available at the Bodleian Library. A capture of the folio by the Selene Photometric Stereo System, which records the surface texture of manuscripts for the ARCHiOx project, showed no visible traces that might further facilitate textual reconstruction. But the text of Hebban olla vogala and its Latin translation can still be read with some certainty as:

                              ;;;;;;; quid expectamus nunc       ·

Abent omnes uolucres nidos inceptos nisi ego & tu

Hebban olla uogala nestas hagunnan hinase hic

anda thu uuat unb\i/[ạ]dan uue nu ·


Largely, this reading validates that of Caron (1963), except for the form anda instead of enda (‘and’), as both the shaft and the bow of the initial a are still faintly visible. It also confirms that unbidan (‘wait’), in which the expunctuation of the first a was rejected by Louwen (2009, p. 67), was seemingly carefully corrected, and that both the Latin nunc (‘now’) and the vernacular (h)ic (‘I’), read as “nu” and “hi” by Louwen (2009, pp. 65–66), were originally written out. Finally, this also establishes that the vernacular text was given an exact Latin translation, which is most likely later since it was crammed into the available space above Hebban olla vogala.

Fig. 2
figure 2

Source: Photograph taken by Kate Fulcher under UV light, filter 500 nm, aperture F8

MS Bodley 340, fol. 169v (detail).

Full size image

Faced with this carefully copied, corrected, and translated pen trial, we may reasonably assume that its linguistic form is quite deliberate. Louwen’s extensive analysis of his reading (2009, p. 78) concludes that Hebban olla vogala is linguistically mixed, combining Dutch, English, and linguistically ambiguous forms. On the one hand, the plural ending in -an of the verb hebban (‘have’) and the forms olla (‘all’) and vogala (‘birds’) are clearly indicative of coastal Dutch (Louwen, 2009, pp. 68–70). Additionally, the scribe’s consistent use of < a > to represent the schwa sound may point to a West Flemish dialect (Cotman & Taeldeman, 2003, pp. 226–228). On the other hand, the distinctly English prefix a- in (h)agunnan (‘begun’), as opposed to bigunnan, has long been recognised as the strongest evidence for Anglicisms in the text.Footnote 8 Similarly, the subjunctive ending in -e of hinase (‘except’) suggests English rather than Dutch, in which the expected ending would be -i (Louwen, 2009, p. 72). As a result, De Grauwe’s conclusion (2004, p. 53) that Hebban olla vogala represents “Kents met West-Vlaams substraat” (‘Kentish with a West Flemish substrate’) does not seem unjustified. At the same time, however, these classifications rely on the slightly simplistic assumption of a clear distinction between dialects of Old English and Old Dutch.

As Mark Faulkner has recently pointed out (2022, pp. 32–64), the linguistic study of late Old English and early Middle English is complicated by the traditional gap between corpora and dictionaries of Old and Middle English, which can mask linguistic diversity in the wake of the gradual collapse of the late West Saxon Schriftsprache. A look at contemporary linguistic evidence accordingly paints a more differentiated picture. This begins with hebban: the form is evidently much more typical of Dutch than it is of English, in which the expected late West Saxon form would be habbað (DOE, s. v. habban). But there is no need to assume, like De Grauwe (2004, p. 46), that it could instead represent the Old English subjunctive plural hæbben, which would significantly alter the meaning of the text. In fact, the indicative plural form hauen is found in the East Midland dialect of the interpolations made c.1121 into the Old English Peterborough Chronicle (Irvine, 2004, p. 31). This suggests that hebban, though not typically Kentish, would have been perfectly intelligible in England. Conversely, Louwen identifies the accusative plural form nestas (‘nests’) as English, since he asserts that Old Dutch lacked the plural ending in -s (2009, p. 70; see also Philippa, 1981, p. 94; but cf. Marynissen, 1994, p. 84). Indeed, while the form is not standard in late West Saxon, in the tenth century, the Northumbrian Lindisfarne Gospels gloss the Latin nidos (‘nests’) with “nestas vel nesto” (London, British Library, Cotton MS Nero D IV, fol. 41v). Such neuter a-stem nouns were the first to take the s-plural in early Middle English, although this development occurred latest in the Southeast (Newman, 1999). At the same time, the s-plural is also a feature of the slightly later West Flemish dialect in the region of Calais, which shows significant English influence (Derolez, 1974, p. 14). As Rolf Bremmer remarks (1989, p. 88), the s-plural in nestas could even be the result of Latin or French influence. English and Dutch in the late eleventh century are a lot more permeable than we tend to assume, which significantly undermines attempts to classify Hebban olla vogala, in whole or in part, as definitively one or the other.

In fact, there seems to be an emphasis on the use of words that can readily be understood in either language. This is perhaps clearest in the case of the word unbidan (‘wait’), which De Grauwe (2008, p. 103) analyses as a composite combining a Dutch prefix and an English verb stem. Louwen (2009, p 76), on the other hand, reconstructs the word umbidan and connects it to the early Middle Dutch ombiden (VMNW, s.v. ontbeiden II). But there is also one recorded instance of the early Middle English vnbiden, which appears instead of the more usual anbiden in a thirteenth-century manuscript, written in an East Midland dialect in Cambridgeshire (London, British Library, MS Arundel 248, fol. 154r; MED, s.v. anbidien, anbīden v.).Footnote 9 The rare form of the word found in Hebban olla vogala can thus still straddle both languages. Most of the text falls into the same category. The form anda (‘and’) not only captures the Dutch ande (VMNW, s.v. ende III) but also has English equivalents. The late seventh-century Kentish Laws of Wihtred, which were copied at Rochester c.1124 in the Textus Roffensis, preserve one instance of the Old English ænde (Rochester, Cathedral Library, A.3.5, fol. 5v; see DOE, s.v. and, ond),Footnote 10 and the Middle English ande also occurs in Dan Michel’s Kentish Aȝenbite of Inwit from 1340 (Gradon, 1965, 1:13; see De Grauwe, 2004, p. 51). Equally, while wat (‘what’) is generally taken as purely Dutch due to the lack of the initial h- expected in Old English (Louwen, 2009, p. 74), the word also occurs as wæt in Old English or wet in the Middle English Aȝenbite of Inwit (DOE, s.v. hwā, hwæt; Gradon, 1965, 1:265; see also De Grauwe, 2004, p. 51). As for (h)ic (‘I’), thu (‘you’), we (‘we’), and nu (‘now’), all of these can evidently be Dutch just as well as English (Louwen, 2009, pp. 72–77). It seems slightly overzealous to explain all this away as the “Ingvaeoonse eigenaardigheeden” (‘Ingvaeonic peculiarities’) of a West Flemish dialect (Schönfeld, 1933, p. 6). The much more attractive hypothesis is that, as Xavier Dekeyser suggests, this pen trial is not only a probatio pennae but also a “probatio linguae” (2007, p. 124), a linguistic experiment. Even so, to characterise this experiment as a partially failed attempt at writing in English would be uncharitable. Instead, the Dutch-speaking scribe of Hebban olla vogala appears to capitalise on the similarity between Old English and Old Dutch to produce a carefully formed text in a Dutch-English Mischsprache which is intelligible in both.

This leaves us with the issue of literary form. The question of the verse form of Hebban olla vogala, or even whether it is verse at all, remains unsettled (Willaert, 2021, pp. 22–23). Peter Dronke (2005) read it as a syllabic poem, which together with its Latin translation forms a bilingual song. But given its irregular syllable count, Hebban olla vogala scans more easily as accentual verse, following a pattern of rhyming half-lines. These occasionally occur in late Old English alliterative verse (Bredehoft, 2005, pp. 79–80), but they are prevalent in late Old High German and early Middle High German poetry (Hoffmann, 1981, pp. 28–63) and still appear in the earliest Middle Dutch poems (Van den Berg, 1983, pp. 155–188). Two half-lines, linked by rhyme or half-rhyme at the end, form a long lineFootnote 11: vogala is thus linked with hagunnan and thu with nu. Furthermore, the same pattern is discernible in the Latin version of the text, which can also be divided into half-lines linked by the consonance between volucres (‘birds’) and inceptos (‘begun’) and the assonance between tu (‘you’) and nunc (‘now’). This gives us two long lines of Dutch-English verse, accompanied by their Latin equivalent, which may be laid out like this:

Hebban olla vogala      nestas hagunnan

hinase hic anda thu      wat unbidan we nu.

Abent omnes volucres      nidos inceptos

nisi ego et tu      quid expectamus nunc.

(All birds have      begun nests,

except for me and you;      what are we waiting for now?)


In this light, Hebban olla vogala emerges more clearly as a careful literary production: a multilingual and translingual poem that moves between different literary traditions and must be seen in all its literary contexts.

The Literary Context in England

While Hebban olla vogala is revered as the beginning of the Dutch literary canon,Footnote 12 its literary function is still uncertain. Since no other contemporaneous Dutch poetry survives, the poem is generally seen in the wider tradition of the European love lyric (Dronke, 2005; Van Oostrom, 2006, pp. 99–107; Willaert, 2021, pp. 18–22), but its monastic context has also encouraged its interpretation as religious verse (Caspers, 2019). Again, however, this is something of a false dichotomy. The function of the text becomes much clearer when considered alongside the other pen trials in MS Bodley 340, fol. 169 (Fig. 1), which have only rarely been edited.Footnote 13 The Dutch-speaking scribe exclusively copies verse, including a Latin antiphonal prayer to St Nicholas and hexameters on music above Hebban olla vogala (see Nix, 1981), as well as the rhyming hymn Rector celi nos exaudi (‘ruler of heaven, hear us’) below. But to the left of the vernacular poem, he also adds the words “Ageiam precor mearum comes inremota rerum” (‘come now, I pray, intimate partner in my affairs’).

These are the opening lines of the fifth-century Versus ad coniugem attributed to Prosper of Aquitaine (see Santelia, 2009), which provides an interesting parallel to Hebban olla vogala. Not only is the Versus a devotional poem taking the form of an address to a lover, which shows that the two divergent interpretations of Hebban olla vogala are very compatible. The Versus was also widely used for linguistic and metrical instruction in England across the Norman Conquest (Ericksen, 2019, p. 100). In fact, one likely context in which our scribe may have encountered the lines is Bede’s De arte metrica, which uses them to illustrate Latin anacreontic verse (Sizoo, 1956/57, p. 124; Willaert, 2021, pp. 16–17). This indicates that form may be as important as content to the vernacular poem’s literary function. Since the Dutch-speaking scribe shows a clear interest in Latin verse forms, the inclusion of Hebban olla vogala among his pen trials suggests that it is also a technical exercise. Caron (1963) speculated that it is an etymological game, striving to create a word-for-word and even sound equivalence with its Latin translation. We may go even further: it appears to be a literary experiment to produce a poem readable not only in Dutch but also in English, and which can seamlessly be translated into a Latin poem using the same verse form.

Caron (1963, p. 257) was also right to ask why the poem should be bilingual, although the question should not only be why Hebban olla vogala was translated into Latin. The Dutch-speaking scribe probably translated the poem for his fellow monks from Normandy, who may have spoken neither English nor Dutch—successfully, it seems, since a different hand copied out the first three words of the Latin further down on the same folio. It is, however, also worth asking what prompted a monk from the Low Countries to copy out a poem in his Dutch-English Mischsprache in the first place. After all, the other Continental hands among the pen trials in MS Bodley 340 and its sister volume MS Bodley 342, which include not only Norman but also German and Italian hands (Kwakkel, 2013), write only in Latin. The most likely answer is that Dutch was also a viable means of communication with English-speakers outside Rochester’s cathedral priory. We consequently need to revise our assumptions about the status of English after the Norman Conquest. Whereas “French and Latin are more easily imagined as ‘international’ or transcultural languages”, it is conventionally thought that “English is only spoken in England” (Dolmans, 2020, p. 28). Broadly, of course, this is true. Yet we should not assume that English was incomprehensible to all post-Conquest arrivals (see Kwakkel, 2005, pp. 18–19). English was probably also the most accessible means of communication for new arrivals from the Low Countries.

In this way, Hebban olla vogala points us towards an easily overlooked movement of cultural exchange. There have been extensive studies of English-Norse mutual intelligibility and bilingualism in early medieval England (Townend, 2000; 2002). But the case for English-Dutch mutual intelligibility in the eleventh century is possibly even stronger. The similarity of the two languages is notably suggested by the Flemish monk and hagiographer Goscelin of Saint-Bertin, one of several Flemish hagiographers in England whose apparent fluency in Old English was an important skill for hagiographic research (Van Houts, 2006b, pp. 121–122). In his Liber confortatorius, written c.1083 and thus exactly contemporaneous with Hebban olla vogala, Goscelin states:

Multe autem gentes uni lingue subiacere noscuntur […] ut Gallis Franci, Alobroges, Aluerni, Beturici, Normanni, Cenomanni, Andecaui, Pictaui, VVascones, Barzellones; ut Theuthonicis Alemanni, Baioarii, Orientales Franci, Dani, Saxones, Thuringi, Lotharingi, Braibandi, Flandri atque Angli (Talbot, 1955, pp. 86–87)

(‘Yet we know that many peoples are subsumed under one language, as under the Gauls the Franks, the Allobroges, the Arverni, the Bituriges, the Normans, the Cenomani, the Angevins, the Poitevins, the Gascons, the Catalans; as under the Teutons the Alemanni, the Bavarians, the East Franks, the Danes, the Saxons, the Thuringians, the Lotharingians, the Brabantians, the Flemish and the English.’)


Clearly, for Goscelin, Flemish and English are part of the same Germanic dialect continuum, which parallels the neighbouring Romance dialect continuum. That is not to say that the transition was generally seamless, as in Hebban olla vogala. A counterexample occurs in the Historia translationis sanctae Lewinnae, written c.1060 by Drogo of Saint-Winoc. Drogo mentions that two Flemish monks were unable to read parchments placed around a Southern English saint’s shrine (Sollerius et al., 1727, p. 615): “quia Anglice, uti apud ipsos mos habetur, scriptae erant” (‘because they were written in English as was the custom among these’). Yet while this has been taken to mean that the Flemish monks did not understand English (Defries, 2008, p. 427), the largest barriers to communication in this case were probably scribal habits. Old English was usually written in a vernacular minuscule, whose letter forms differed from the more “international” Caroline minuscule used for Latin from the late ninth century (Bishop, 1971, p. xi; see Roberts, 2015, pp. 38–41). English also used a set of distinctive letters, including < ƿ > (wynn), < ð > (eth), and < þ > (thorn), which in a twelfth-century alphabetic table from Peterborough (London, British Library, MS Stowe 57, fol. 3v) are set apart from the Latin alphabet as “Anglicae litterae” (‘English letters’). In contrast, spoken English, or even English spelt following Latin conventions, would have presented far fewer challenges to a Dutch-speaker.

The high Middle Ages certainly saw frequent exchanges between England and the Low Countries. English exiles often took refuge in Flanders, both before and after the Norman Conquest (Grierson, 1968, pp. 78–85). One of these was Gunhild, sister of Harold Godwinson, who went into exile in Flanders in 1067 and may have left her Old English glossed psalter to St Donatian’s Church in Bruges (Opalińska et al., 2023; Porck, 2024; Van Houts, 2023). Before the Norman Conquest, the later English rebel Hereward was exiled to Flanders,Footnote 14 where his military activities were probably aided by the mutual intelligibility of English and Dutch (Van Houts, 2006a, p. 5). It should then be no surprise that the language of Hebban olla vogala was traditionally located in the West Flemish dialect region, west of Saint-Omer, where English influence was greatest and linguistic differences smallest (Gysseling, 1980, p. 129; Cotman & Taeldeman, 2003, p. 224). In the same way, Kentish was particularly subject to influence from Flanders (Samuels, 1971). Eastern England and Flanders should both be seen as part of a “North Sea zone” with considerable cultural exchange, even if the extent of Flemish influence in England after the Norman Conquest was partly masked by the French names of many post-Conquest settlers (Crick, 2021, pp. 75–77).

There was in fact a considerable Flemish presence in England in the late eleventh and twelfth centuries. In c.1151, the chronicler Alfred of Beverley (Slevin & Lockyer, 2023, p. 9) added to the traditional five peoples of Britain “sexta natio, id est Flandrenses” (‘a sixth nation, that is the Flemings’). While he primarily refers to Flemish settlements in Wales, the Flemish presence was clearly considered “of national significance” (Oksanen, 2012, p. 178). Similarly, Geoffrey of Monmouth’s Historia regum Britanniae, completed c.1135, conspicuously integrates a fictionalised Flemish people into its rewriting of British history (Lumbley, 2023). Yet as Elias Oksanen has shown (2012, p. 220), the Flemish population of England was not consistently set apart as an ethnic group. In part, this may be because, on a linguistic level, they could much more easily blend into the English-speaking population.

This has far-reaching consequences for English literary history and Hebban olla vogala’s place in it. Faulkner argues that, until c.1140, the “premier literary vernacular of England” was still English rather than Anglo-Norman French (2022, p. 91). Hebban olla vogala must be seen as participating in this English literary culture. But the poem also illustrates the difficulty of distinguishing Dutch influence on English due to their close similarity (Chamson, 2014, pp. 282–283). Some linguistic influence is visible in English literature of the late twelfth century. The Ormulum, written in the second half of the twelfth century in Lincolnshire, contains the word scone (‘beautiful’). The work’s distinctive orthographical system, which indicates an initial /sk/ rather than /ʃ/ (Anderson & Britton, 1999, p. 305), sets this apart from the early Middle English shene (MED, s.v. shẹ̄ne adj.). Accordingly, the Ormulum’s most recent editors link scone to either Old Frisian or Middle Dutch (Johannesson & Cooper, 2023, p. 635). In fact, it likely derives from the early Middle Dutch scone (VMNW, s.v. scone V), as in Old Frisian the expected form would be skene (Bremmer, 2004, p. 133; Boutkan & Siebinga, 2005, p. 347). The similarity between Dutch and English, however, may also conceal broader literary influence. If, as Elisabeth van Houts argues, “an English person in 1066 might have found it easier to converse with a Flemish Dutch speaker than with a French person” (2006a, p. 5), they would have also found it easy to exchange literary works. Hebban olla vogala, which is carefully constructed to become part of English as well as Dutch literature, is only the most visible element of what could have been a much broader current of literary exchange.

The poem’s form represents the most tangible aspect of its place in English literary history. Its structural reliance on rhyme is distinctly Continental: High and Low German poetry saw a shift from alliteration towards rhyme as early as the ninth century, long before similar developments in Old English (von See, 1967, pp. 72–75; Hoffmann, 1981, p. 28). In this context, it is worth noting that Hebban olla vogala has repeatedly been compared to Latin love lyrics from German-speaking areas (Dronke, 2005; Van Oostrom, 2006, pp. 99–102), such as the eleventh-century lyric Cantant omnes volucres (Sankt Florian, Stiftsbibliothek, XI.58, fol. 83v) and the Cambridge Songs (Cambridge, University Library, Gg.5.35). This latter collection of tenth-century Latin verse also contains two rhyming macaronic Latin-German poems, but it was copied in mid-eleventh-century England and kept at St Augustine’s Abbey, Canterbury (Ziolkowski, 2020, p. xxxi). Jan Ziolkowski argues that concomitant German influences on English literature were largely stymied by the Norman Conquest, which created “a cultural axis that ran in a very different direction” (2020, p. xxxii). The presence of Hebban olla vogala in MS Bodley 340, however, indicates that Dutch influence was, if anything, invigorated.

Although no similar manuscript evidence survives, Hebban olla vogala should therefore at least be taken into account when charting the development of English literary forms. Recent histories of English alliterative metre trace a continuity across the Norman Conquest, which includes partly rhymed types of late Old English poetry.Footnote 15 Rhyme is a recurring feature of the post-classical metre of some late Old English chronicle verse (Bredehoft, 2001, pp. 96–99). Rhyme as the main structural feature in English poetry, however, only seems to have become established in the aftermath of the Norman Conquest (McKie, 1997, p. 829; Stanley, 1988, p. 37). The best example is the poem on the death of William the Conqueror included in the Peterborough Chronicle’s entry for 1086. Its structural reliance on rhyme has been called “idiosyncratic” (Bredehoft, 2004, p. 95), but the verse form of this condemnation of William’s despotism is in part very similar to that of the contemporaneous Hebban olla vogala:

ac hi moston mid ealle      þes cynges wille folgian

gif hi woldon libban      oððe land habban,

land oððe eahta      oððe wel his sehta. (Irvine, 2004, p. 98)

(‘But they had, in all things,      to follow the king’s will

if they wished to live      or hold land,

land or property      or indeed his friendship.’)

Seth Lerer (1999, p. 18) argues that this is “the first poem in rhymed couplets in the English language”, which reinforces its narrative of oppression by using what he sees as a verse form derived from Norman French rhyming couplets. But the poem on William’s death (Irvine, 2004, p. 97) explicitly chronicles the king’s oppression of both “rice men” (‘rich men’) and “earme men” (‘poor men’), without differentiating between William’s English, Norman, or even Flemish subjects. In the same way, Norman French was hardly the only potential source of influence on late Old English literature. After all, this is precisely the verse form we see slotting itself into English in Hebban olla vogala – which could thus equally be called “the first poem in rhymed couplets in the English language”. A similar pattern of accentual rhyming half-lines, as opposed to French-derived syllabic verse, can then still be identified as late as the thirteenth century in the early Middle English romance King Horn (Putter, 2023, p. 133). Among the various rhyming forms available in post-Conquest England, the possibility of Dutch literary influence should consequently not be entirely discounted.

In any case, this highlights that late Old English and early Middle English literature is part of a more multilingual literary network than commonly acknowledged. Another case in point is Laȝamon’s Brut. Composed c.1200 in Areley Kings in Worcestershire, it translates Wace’s Norman adaptation of Geoffrey’s Historia regum Britanniae into early Middle English. Laȝamon’s verse form shifts between alliterating and rhyming half-lines and has therefore often been described as a blend of different traditions (Salter, 1988, p. 48; Pascual, 2017, p. 257). Yet due to the perceived insularity of English, Laȝamon’s choice to write in English is still frequently seen as a mark of linguistic nationalism directed at local audiences (see Le Saux, 1989, p. 227). This changes once we consider the permeability of English and Dutch and the resulting transnational reach of English. The prologue to the Caligula text of Laȝamon’s Brut accordingly presents a highly capacious vision of English history. Whereas Geoffrey and Wace introduce the ancient Britons as the first settlers of England, Laȝamon announces:

þet he wolde of Engle; þa æðelæn tellen.

wat heo ihoten weoren; ⁊ wonene heo comen.

þa Englene londe; ærest ahten. (Brook & Leslie, 1963–78, ll. 7–9)

(‘That he wished to tell of the noble English,

What they were called and whence they came,

Who first held England.’)


Laȝamon thus redefines Englishness as based on possession of English land—which gives all those holding land in England a stake in English history. Indeed, as Rosamund Allen points out (1998, p. 15), the manor of Martley, within which Areley Kings lay, was from 1204 held by William de Frise, a Flemish or Frisian knight of King John, making William and his household part of Laȝamon’s potential audiences.

The permeability of English and Dutch in the high Middle Ages, however, should not be exaggerated, and nor should possible Dutch influence. Laȝamon’s vocabulary appears deliberately archaising (Stanley, 1969). Speculating that William de Frise may have been Laȝamon’s patron, Allen (1998, p. 16) then suggests that archaic English might have been particularly easy to understand for a knight from the Low Countries. Yet Allen relies on a reading in the Otho text of Laȝamon’s Brut (Brook & Leslie, 1963–78, l. 3), which states that Laȝamon lived “wid þan gode cniþte” (‘with the good knight’), and which due to Otho’s textual revisions is highly questionable (Roberts, 2013, p. 454). It is of course trivially true that some of Laȝamon’s archaic or poetic terms have transparent cognates in Dutch or Frisian. For example, the archaic word hæleþ (‘man’, ‘warrior’) (see Wyld, 1930: 11) corresponds to the Middle Dutch helet (MNW, s.v. helet). But there are many elements of Laȝamon’s specialised diction that have no such cognates, including much of the poetic vocabulary typical of English alliterative verse, such as the other terms for ‘man’ beorn, segg, and rink (see Wyld, 1930, pp. 4, 18; Turville-Petre, 1977, p. 79). Failing better evidence, literary interactions between Dutch and English in the wake of Hebban olla vogala remain a potentially underlying current rather than a clearly identifiable factor.

Conclusion

Hebban olla vogala emerges not only as part of Dutch literary history, but also as early evidence of the interactions between Dutch and English literature. In exploiting the mutual intelligibility of Dutch and English, it points us towards what may have been a much broader current of Dutch literary influence in post-Conquest England. But we must also recognise that this remains difficult to pin down, and while the possibility may seem tempting, not everything that looks like evidence of Dutch influence necessarily is. This notably applies to another scribal addition in MS Bodley 340. On fol. 1r, several Old English words at the beginning of one of Ælfric’s homilies were glossed in the thirteenth century. Most of these glosses are in Latin, but two are vernacular, the first of which glosses the Old English “casere” (‘emperor’) as “keysere”. Based on the gloss’s presence in the same manuscript as Hebban olla vogala, Kwakkel interprets this as Middle Dutch (2005, p. 8),Footnote 16 but of course keysere is also an unremarkable Middle English form (MED, s.v. caiser n.). Since the second vernacular addition glosses the Old English “ceastre” (‘town’) as the unambiguously Middle English “chestre” (MED, s.v. chestre n.), the balance of probability suggests that both terms are Middle English.

Nevertheless, Hebban olla vogala still provides an important counterpoint to the focus on the confluence of English, French and Latin that dominates the study of twelfth-century English literature (see Short, 1991; Ashe, 2022, pp. 78–85). It indicates at least the ease with which Dutch literary traditions could also find a place in England. This is, by nature, hard to trace on a larger scale. Kwakkel notes that none of the non-Norman hands responsible for pen trials in MS Bodley 340 occur in other Rochester manuscripts—and he suggests that, by adopting the standardised “prickly” script of Rochester when not trying their pens, they may be “hiding in plain sight” (2013, p. 254). Equally, Emil Chamson posits that, due to the large degree of similarity between the two languages, Dutch linguistic influence on English may remain “hidden in plain sight” (2014, p. 300). The same could then be said of Dutch influence on English literature: the very similarity that makes interaction and influence possible is also what makes it so difficult to identify.