Uniquely extant in Hauksbók (Hb, AM 544 4°), Merlínusspá ‘The Prophecies of Merlin’ (GunnLeif Merl) is preserved as two separate poems (termed here Merl I and Merl II) that form part of Breta saga ‘The Saga of the British’ (Bret), a rendering of De gestis Britonum ‘Concerning the deeds of the Britons’ (DGB: formerly known as Historia regum Britanniae ‘History of the kings of Britain’) (Halvorsen 1959, 22-3; cf. Lönnroth 1965, 83). The present Introduction is pertinent to both parts of Merl and their relationship is discussed below. The author of Merl is identified in Hb as follows (Bret 1848-9, II, 12, here normalised): Hér eptir hefir Guðlaugr munkr ort kvæði, þat er heitir Merlínusspá ‘In accordance with this, Guðlaugr the monk has composed a poem called Merlínusspá’. The reading Guðlaugr found in Hb appears to result from later over-writing of original Gunnlaugr (Bret 1848-9, II, 13 n. 9). The section of Hb where Merl occurs was written by Haukr Erlendsson himself in the first decade of the fourteenth century (Stefán Karlsson 1964b; cf. Johansson 2005, 111); the text of Merl begins on fol. 49r and ends on fol. 53r. The AM 573 4° redaction of Bret corroborates the ascription of Merl to Gunnlaugr but omits the text, observing (Bret 1848-9, II, 13 n. 11): kunna margir menn þat kvæði ‘many people know that poem’, where kunna is assumed to mean ‘know by heart’ (Jakob Benediktsson 1966, 556). The attribution to Gunnlaugr has never been seriously questioned in scholarship (though for a suggestion, on insufficient evidence, that part of the text was composed by an imitator of Gunnlaugr rather than Gunnlaugr himself, see Horst 2006).
Gunnlaugr is thought to have completed Merl c. 1200 (for somewhat earlier suggested datings see Leach 1921, 138; Sveinbjörn Rafnsson 1999, 391). The poem represents an Icelandic rendering of a prose narrative by Geoffrey of Monmouth entitled the Prophetiae Merlini ‘Prophecies of Merlin’, which is now best known to us as a section of the DGB. The DGB probably assumed its finally published and transmitted form at some point in late 1137 or in 1138 (cf. Tatlock 1950, 433-7; Gransden 1974, 201; Wright 1985, ix-xvi; Reeve and Wright 2007, vii). The text-within-a-text constituted by the Prophecies occupies DGB VII chs 3 and 4 in the edition of Griscom (1929), corresponding to chs 111-17 of the edition of Faral (1929) and of Reeve and Wright (2007, 142-59; the page and line numbers of the latter edition are used in the ensuing discussion and Notes). The Prophecies are thought to have been separately issued by Geoffrey in 1135 (Wille 2015, iv).
In the narrative of DGB, the Prophecies are prompted by the attempts of Vortigernus, king of the Britons, to build a mountain-top fortification against the invading Saxons. The builders find that the foundations they build one day have mysteriously disappeared by the next. The king seeks the advice of wise men who counsel him to have the mortar mixed with the blood of a child without a father (iuuenem sine patre: Reeve and Wright 2007, 137). Questing for such a child, the king’s envoys discover Merlin, whose father was an incubus. Rather than be killed as the treacherous wise men propose, Merlin discloses the presence of two serpents that are responsible for undermining the foundations and goes on to reveal at great length what they portend for the Britons. These prophecies cover aspects of British history and legend starting with the wars between the Anglo-Saxons and the Britons (Prophecies 1 to 8, as numbered in Wright 1988), continuing with the Norman Conquest and the Plantagenet kings (9 to 12) and finally projecting onward to events that lay in the future from Geoffrey’s viewpoint so as to conclude with the end of the world (13 to 74). The last historical event definitely verifiable in the Prophecies is the drowning of Henry I’s heir apparent and some of his other offspring in the year 1120 (Tatlock 1950, 403). Thereafter the events described bear no clear relation to known historical events, though see Merl I 53 and 54, Notes to [All]. Geoffrey’s distinctive use of an intermittent allegory, where rulers are represented by one or more animal figures (lion, lynx, boar, fox, wolf, serpent or combinations thereof), renders the content of the Prophecies highly cryptic and enigmatic (Taylor 1911, 4, 11). Many readers will wholeheartedly concur with Tatlock (1950, 416), who describes the Prophecies as ‘imaginative to the point of lunacy’ – ‘much of them it is hard to believe ever had any intelligible meaning for anyone’. While such strictures are partly justifiable, and will have been concurred in by at least a few medieval readers, the importance of the prophecies of Merlin and the numerous vaticinatory texts composed under its influence in the workings of society, both English and European, right down to the early modern period, is well documented (see e.g. Aurell 2007, 391; Dobin 1990, 25, 51; Curley 1984, 335-6; Fraiolo 1982). The reign of Henry II saw particularly intense scrutiny of the text with an eye to signs and portents (see e.g. Warren 1978, 25). The redactor of the AM 573 4° version of Bret notes this fact (Bret 1848-9, II, 13): Sv spá hefir oft síðan af hinum uitrvztvm mönnum á Einglandi rannsökuð uerið ‘That prophecy has often since been inquired into by the wisest men in England’. Gunnlaugr enjoins readers of his adaptation to scrutinise it in the same light (see Merl I 101). Towards this end, perhaps, he appears to put the credentials of the Prophecies on their best footing by vouching for Merlin’s Christian status and by adducing only the Old Testament prophets Daniel and David as analogues rather than the Sibyls and other non-Scriptural figures adduced by some European commentators.
Symptomatic of its extraordinary popularity and rapid dissemination, the DGB survives in over 200 mss, of which about fifty belong to the twelfth century; in addition, over eighty mss of the Prophecies, separate from DGB, are extant (Veysseyre and Wille 2008, 93). It is not known how and in what form Geoffrey’s work reached Iceland. A Lincoln connection has been tentatively suggested by Leach (1921, 138-9) on circumstantial grounds. The bishop of Lincoln, Alexander the Magnificent (1123-43), as dedicatee of Geoffrey’s separate publication of the Prophecies, received the first copy of it from Geoffrey (Wright 1984, 1) and it could have been copied and made available to readers there, among them members of the celebrated cathedral school at Lincoln. This school had Icelanders among its international band of students, notably, around 1160, Þorlákr Þorhallsson, later bishop at Skálholt 1178-93, and his successor in the see Páll Jónsson (Orri Vésteinsson 2000b, 154). The intense interest in the Prophecies in England during the reign of Henry II, prompted not least by the king’s antagonism toward Thomas Becket, coincided with an era of conflict over leadership policies and the limits of ecclesiastical jurisdiction in Iceland (cf. Helgi Þorláksson 2005, 146; Guðrún P. Helgadóttir 1987, xii-xiv). Prominent in the debate was Guðmundr Arason, a celebrated priest and later bishop of Hólar (1203-37). Close relations with this see were maintained by Gunnlaugr’s institution of Þingeyrar, near the shores of Húnaflói (Turville-Petre 1953, 135). The most noteworthy honour accorded to Guðmundr prior to his election as bishop was the procession led by Karl, abbot of Þingeyrar, and Gunnlaugr to greet him on his visit to the monastery in 1199 (Ciklamini 2004, 66). Gunnlaugr for his part was regarded in his own time as a man of singular Latin learning (ÍF 15, cclxxxiii). He is known to have been entrusted by Guðmundr with commissions for writings pertaining to S. Þorlákr and Bishop Jón Ǫgmundarson, both in Latin (Turville-Petre 1953, 198-9). More than this, Guðmundr actively provided some of the content by sending Gunnlaugr accounts of portents and miracles relating to S. Þorlákr. Given this relationship between Guðmundr and Gunnlaugr, the former might well have been responsible for commissioning Gunnlaugr’s translation of the Prophecies. As Crick (2011, 76) notes, ‘Merlin was a matter for bishops and popes’ and Guðmundr might well have aspired to an episcopal role long before his election.
Alternative possibilities exist as to the character of Gunnlaugr’s source ms.:
1) He might have based himself on a ms. of DGB. This assumption would be favoured by the fact that, codicologically speaking, the unique attestation of Merl occurs within the Hb redaction of Bret, the Icelandic prose translation of the DGB, and is placed in a context corresponding to that of the prophecies in DGB. In the Hb redaction the poem is preceded by a brief paragraph mentioning the poet Gunnlaugr, which, although apparently an interpolation, in fact does not preclude the presence of a non-Gunnlaugr text of the prophecy in Haukr’s source. In the other extant redaction of Bret, that of AM 573 4° (fol. 45r), the situation is similar. An advantage of this hypothesis is that it obviates the need to posit two importations of Geoffrey material to Iceland: first, a separate copy of the Prophecies for Gunnlaugr to translate; second, at some unknown time, a copy of the complete DGB which, whether directly or through intermediaries, yielded the extant redactions of Bret. On the other hand, this application of Ockham’s razor is far from conclusive, since while we know that large quantities of learned books were brought to Iceland in the relevant period we have very little by way of precise inventories from which to gauge patterns of acquisition. Moreover, as we shall see later, it would be difficult to prove that Gunnlaugr personally had access to the full DGB.
2) Gunnlaugr might have worked from a separate ms. of the Prophecies. Within this scenario there are again two possibilities to be distinguished. He might have used a copy of the Prophecies that derived from the full DGB. Alternatively, he might have had recourse to a copy of the Libellus Merlini ‘The Booklet of Merlin’, an early version of the Prophecies that appears, as noted, to have been separately issued by 1135. Evidence for its separateness takes three forms. Ordericus Vitalis quotes the equivalent of DGB chs 113:72-115:108 in his Historia ecclesiastica ‘Ecclesiastical History’ (1135), stating that they emanate from a Merlini libellus (Reeve and Wright 2007, viii-ix; contrast Tatlock 1950, 418-21). In some mss the Prophecies are separately dedicated to Bishop Alexander. Finally, Geoffrey states that he accessed the Prophecies in a British-language source distinct from the one purportedly used for DGB (Reeve and Wright 2007, 142-3). While this claim, as applied to the Prophecies as a whole, is no doubt spurious, he may have had access to a Brittonic source or sources for at least some of his material (Faletra 2012, 306, 309).
In considering the rival claims of the two possible source types within scenario 2, we can refer to the study of manuscript affiliations carried out by Michael D. Reeve (in Reeve and Wright 2007). Reeve posits a separate text of the Prophecies, called Π, which he equates with the Libellus. As he recognises, not all complete mss of the Prophecies descend from this text: some appear to have had an incomplete ancestor that was later supplemented from DGB (Reeve and Wright 2007, xxix). In order to determine which separate copies of the Prophecies descend from Π and were not extracted from DGB, Reeve singles out two of Ordericus’ readings, 113:74 ipsius ‘of the same’ for albi draconis ‘of the white serpent’ and 114:92 translateralibus ‘adjacent peoples’ for collateralibus ‘adjacent peoples’ (Reeve and Wright 2007, xxix). To judge from Merl I 47/6: snáks ins hvíta ‘of the white snake’, Gunnlaugr’s source ms. would have contained albi draconis ‘of the white dragon’. The status of the other decisive reading, at 114:92, cannot be determined, since it is not represented in the extant text of Merl. This diagnostic, taken at face value, would indicate that Gunnlaugr’s source ms. belonged not with Π but with the Ω group, where Ω = a no longer extant ms. of DGB that is posited as the common source of most of the non-Π ms. classes (Reeve and Wright 2007, l). That might well be correct, but the reservation must be made that Gunnlaugr so often uses heiti and kennings instead of pronominal forms of reference that this reading falls short of proof that he is here following a ms. of the Ω class. The same applies to the other readings characteristic of Ω that we find apparently reflected in Merl. These include septem ‘seven’ for octo ‘eight’ in ch. 112.52-3 and tria ‘three’ for quattuor ‘four’ in ch. 116.188 (respectively Merl I 33/3 and II 27/1); numerals are a frequent source of error in the transmission of DGB and coincident error is always possible. Additionally, a sentence at 116:174 which is omitted by the Ω group but represented in Π (and hence in Ordericus) does not appear in Merl (see Merl II 19 Note to [All]); once again, however, such silence might be due to coincident error, since on his own admission Gunnlaugr omits some tracts of text and moreover losses of portions of the text of Merl in transmission subsequent to him are likely to have occurred. For further readings that might be interpreted as indicating the detailed affiliations of Gunnlaugr’s copy-text see Merl I 28, I 29, I 41, I 45, I 54, Notes to [All], Merl II 25 and II 48, Notes to [All]. In sum, the textual evidence, taken in isolation, is scarcely conclusive as to which class of mss Gunnlaugr’s source belonged to (cf. Tétrel 2010, 496). At the same time, the weight of evidence points us in the direction of the Ω group.
In deciding whether his copy-text confined itself to the Prophecies, included the whole of DGB or lay somewhere in between these extremes, we should note that Gunnlaugr appears to have had some knowledge of the contents of certain parts of DGB outside the Prophecies, notably Books IX-XI but also Books V and VI (cf. J. S. Eysteinsson 1953-7). This does not necessarily prove, however, that he had access to DGB per se. From the time of their first publication the Prophecies were immediately copied, glossed, annotated and translated, sometimes alongside the complete DGB and sometimes separately. The Prophecies are one of a very few secular texts of the twelfth century to have become the subject of commentaries written by contemporary scholars and many text-plus-commentary compilations on the Prophecies were produced before the end of the twelfth century. Instances are the twelfth-century mss Lincoln Cathedral Ms. A. 46 (Taylor 1911, 87-8; Crick 1991, 85-7) and Paris BN Lat. 14465, the latter of which adds the chapters of DGB that precede the Prophecies from 105.489 to the end of 108 so as to explain their historical background (Reeve and Wright 2007, xxx). Most of the commentaries so far studied contain a digest of information about Arthur (ultimately stemming from DGB IX-XI). Gunnlaugr could have obtained his supplementary material from such a source. Gunnlaugr seems likely, on other evidence, to have availed himself of a copy with annotation or commentary of some kind: see Merl I 9/5 and I 56 Notes to [All]. In the present state of scholarship, with not all commentaries and annotated versions edited, it is not feasible to reach a firm conclusion on this question. Complicating the matter is the lack of scholarly agreement as to whether the translation of DGB into Bret preceded that of Merl (Turville-Petre 1953, 202), or vice versa (Hb 1892-6, cxi; Paasche 1957, 323); in its extant shape Bret is a much altered version of the original translation (Kalinke 2009, 218; Tétrel 2010, 493-5) and offers no clear indications as to the floruit or identity of the translator (Würth 1998, 56).
Aside from DGB and commentaries on this work, Gunnlaugr appears to have made sporadic use of a few other sources. He refers on occasion to Bede’s Historia Ecclesiastica ‘Ecclesiastical History’ (HE; on Gunnlaugr’s knowledge of Bede, cf. LH II, 402; J. S. Eysteinsson 1953-7, 102, 110), to Henry of Huntingdon’s Historia Anglorum ‘History of the English’ (HA), published in different versions from c. 1131 down to c. 1155 (HA 1996, lxx-lxxvii; on knowledge of Henry in twelfth-century Denmark see Gelting 2007, 106), and to William of Malmesbury’s Gesta regum Anglorum ‘Deeds of the kings of the English’, published c. 1125 (Mynors et al. 1998-9, II, xxiv; on Gunnlaugr’s possible access to William, cf. J. S. Eysteinsson 1953-7, 102, 110). On these possible references see Merl I 2, I 4, I 5 and I 50, Notes to [All], and Note to I 49/4. Additionally, Merl seems to reflect an awareness of contemporary beast epic. This emerging genre was available from the mid-twelfth century in the shape of the Ysengrimus, a Latin epic poem that featured lengthy dialogue passages. It prompted the production of vernacular texts such as the Roman de Renart, which contains a series of short episodes centred on the fox’s hoodwinking of various other animals (Mann 2009, 17-19). These texts proliferate between 1150 and 1200 (Mann 1987, 3) and some of their distinctive features appear in Merl II 37-45, the passage that deals with Geoffrey’s fox king. For Gunnlaugr’s knowledge of vernacular Old Norse literature see below.
So far as we can judge from the extant text of Merl, Gunnlaugr’s translation of the Prophecies per se follows Geoffrey’s text with remarkable closeness and accuracy (cf. LH II, 173). This appears to arise from the nature of the text: as prophetic material, Merlin’s vaticination was treated as tantamount to sacred, and scribes and translators seem to have taken care to preserve the precise wording or sense whenever possible (cf. Blacker 2005, 10). To facilitate comparison, this edition cites both text and translation from Reeve and Wright (2007) for all relevant passages; in those instances where their translation is somewhat free, masking the accuracy of the rendition in Merl, a more literal version has been substituted (marked by the use of ‘cf.’ before the page reference to Reeve and Wright). It must be borne in mind, however, that a definitive edition of DGB, using all extant mss, has yet to be attempted; indeed, even with a complete account of the paradosis the exact wording of Gunnlaugr’s source text is likely to remain indeterminate. The same difficulty applies to any attempt to show that his translation has value for the textual criticism of DGB (for an investigation of this kind into the prose of Bret, see van Hamel 1936), but see Note to Merl I 54/4 for a possible instance of Gunnlaugr’s text reflecting an otherwise unknown superior reading. Exceptions to Gunnlaugr’s fidelity to DGB consist of material added from the other sources noted above, along with material he freely composed, comprising introductions and conclusions to each Part and two matched battle descriptions, each placed approximately centrally within its Part. Some of the apocalyptic material in Merlin’s prophecy is supplanted by a brief homily attacking vanity and urging contempt of the world (Merl II 52-7): Gunnlaugr expressly states in Merl II 50 that he will omit many prophecies on account of their length and distasteful content. In some cases, however, it is difficult, as noted above, to decide whether the absence of these materials reflects deliberate omission on Gunnlaugr’s part or casual loss in subsequent transmission.
Gunnlaugr chose verse as the medium for his translation, and specifically the fornyrðislag metre, varied with use of kviðuháttr at the close of Merl II. The choice of verse rather than prose may have been prompted by a persisting tradition, fostered by William of Newburgh among other commentators (Curley 1982, 221-2), that Geoffrey had a vernacular source in poetic form. The choice of fornyrðislag rather than other metres may be owing to the existence of several well-known Icelandic poems of prophecy in that verse-form, notably Vsp but also Gríp and Hyndl. For a systematic study of the relationship between Merl and Vsp, see Horst (2010); for the view that Vsp was based on Merl, instead of the other way around, as has been the standard assumption in scholarship and in this edition, see Sveinbjörn Rafnsson (1999). It happens that both Merl and Vsp occur in Hb; see Simek (1991) and Johansson (2005) for opinions on the textual relationship between them in that ms. On the other hand, Vsp should not be seen as Gunnlaugr’s unique model (cf. Horst 2010); verbal parallels have been detected not merely between Merl and Vsp but also between Merl and such other eddic poems as the Helgakviða sequence and Gríp: see Notes to Merl I 10/6, I 66/3, I 77/8, Merl II 2/7-8, II 24/1, II 35/7, II 36/6, II 51/6, II 54/1, 3, II 59/2 and II 61/1. (For possible allusions to Hávm see Notes to Merl I 103/1-2 and II 24/4.) These proposed stylistic parallels have been explained as due to Gunnlaugr’s direct borrowing from specific poems (de Vries 1964-7, II, 75-6), but, given the broadly generic nature of the verbal resemblances, most are better explained on the basis that he was simply using a general eddic style such as we also see in the early twelfth-century Gísl MagnkvII and the mid-twelfth-century Ív SigII; for a verbal parallel with Gísl Magnkv see Note to Merl I 67/9. The occasional use of ten- and twelve-line stanzas rather than the standard eight is also best explained as resulting from a general awareness of eddic style rather than indebtedness to any one specific poem. It has been claimed (Turville-Petre 1953, 200) that, as used by Gunnlaugr, the fornyrðislag metre is regularised more than is typical of poems in this verse-form in the Poetic Edda, and this observation, if tenable, suggests his independent command of the medium. Complementarily, Gunnlaugr incorporates some skaldic stylisms, especially in the expanded battle descriptions, which feature an array of heiti and kennings on martial themes that considerably surpasses the incidence of such features to be found in more strictly eddic poetry (LH II, 173-4; de Vries 1964-7, II, 76). Here once again it is easier to see a general resemblance to twelfth- and thirteenth-century skaldic style, as represented by such works as RvHbreiðm HlIII, SnSt HtIII and Anon Krm (Ragn), than an indebtedness to specific poems. As to possible influence of Merl on other poems, the most salient candidate is Ht: the length of Merl I, at 103 stanzas, is closely comparable with that of Ht, at 102 stanzas, and commonalities of certain distinctive kennings for body parts between the two works (Guðrún Nordal 2001, 242, 308) might further indicate some emulation of Gunnlaugr on Snorri’s part. Another important element in the poem is the characteristic vocabulary of homiletic and related literature (see e.g. Notes to Merl I 54/4, I 87/1, II 53/1-4, II 55/8, II 65/1).
As to structure, Merl is made up of two linked poems. The Part designated Merl I in this edition contains Geoffrey’s chs 105.494 to 115.146 (inclusive), covering events from the lead-up to Merlin’s utterances down to the massacre at London narrated in Prophecy 30. The Part designated Merl II in this edition continues from chs 116.1 to 117.304, covering events from the emergence of the three springs at Winchester in Prophecy 31 down to the end of the world concluded in Prophecy 74. Each Part is complete as an independent poem, having its own introduction and conclusion (Hb 1892-6, cxi). In Merl I the introduction consists of nine stanzas that provide information on the historical background, including Merlin, and the conclusion consists of Gunnlaugr’s freely composed apologia for Geoffrey’s allegorical method. In Merl II the introduction consists of four stanzas of explanation about Merlin and the nature of the ensuing poem and the conclusion consists of Gunnlaugr’s freely composed peroration. Of the two Parts, Merl I is the more informative on the historical background and on Merlin, whereas Merl II carries the more elaborate conclusion and peroration. Linking these two Parts are stanzas I 93 and 94, where Gunnlaugr speaks of further prophecies that he has evidently already rendered and means to issue as a separate installment (Merl II). When he says that he has compiled ‘some’ of the prophecies (I 93/7-8), this should be seen not as indicating that the preparation of Part II is still in progress but rather that other prophecies have been deliberately excluded from the project, as he duly states in II 50 (and cf. II 62). From the internal evidence, then, we can draw the conclusion that Gunnlaugr published his translation in the form of two installments, having first prepared the entire poem. Even so, the renewed introduction of Merlin at the start of Merl II is odd. Gunnlaugr may simply have felt that his audience needed a quick reminder about Merlin’s role, but the stanzas read more like an introduction de novo for the benefit of an audience that had not previously heard of Merlin. This feature, along with the reverse-order sequencing of the Parts in Hb, on which see below, has prompted suggestions that Gunnlaugr wrote and published the Parts of his translation project in reverse-chronological order (Bret 1848-9, II, 71 n. 8; cf. LH II, 173; de Vries 1964-7, II, 75; Sveinbjörn Rafnsson 1994, 736; Guðrún Nordal 2001, 88). The likelihood of that is reduced by the factors reviewed above, notably the pivotal role played by Merl I 93-4. On the other hand, it is conceivable that Gunnlaugr might have done a pre-publication limited issue of Part II in booklet form. If such were the case, we could suggest speculatively that the natural person to submit it to would have been Guðmundr Arason in his known capacity (noted above) as Gunnlaugr’s patron or commissioner for other works. The motive could have been to solicit vetting and clearance prior to full publication of either Part (Poole 2009, 310-15). The solicitation of prior approval before full publication would be comparable to and perhaps emulative of that pursued by Ari Þorgilsson (1067-1148) in his preparation of Íslb (on Ari, see ÍF 1, 3; Andersson 1985a, 200; cf. LH II, 397) and would have been all the more important when debate raged in England and Europe as to whether Merlin should be seen as Christian or pagan (cf. Crick 2011, 77). The use of booklets is indicated elsewhere in Old Icelandic manuscript production (Lindblad 1954, 326) and there is abundant parallel evidence from medieval England (Robinson 1978, 231-2; Hanna 1996, 21‑34).
As noted, a very strange feature of the transmission of Merl is that, as presented in Hb and in all editions up until now, the two Parts appear in reverse order, so that the chronologically posterior part of the Prophecies, as sequenced in Geoffrey’s DGB (VII 4 in Griscom 1929), is placed before the chronologically prior part (VII 3 in Griscom 1929) (Hb 1892-6, cxi; Jón Helgason 1960a, xix; Turville-Petre 1953, 200-2). This transposition of Parts could once again be accounted for on the basis that from the outset the translation lived in two separate booklets, each containing one Part, and that the vagaries of distribution and perhaps also factors of individual interest and taste meant that some readers saw Part II ahead of Part I. It would then be possible for a copyist, perhaps Haukr himself, to copy the booklets in the incorrect order, either inadvertently (cf. Merl 2012, 16) or as an artifact of the timing of availability. For the possible use of booklets elsewhere in Hb see the discussion by Gatch (1978, 55) of the two Ælfric sermons excerpted and adapted within it (described by Jón Helgason 1960a, xiii).
In addition to this reversal of Parts, Merl as extant represents a substantially damaged state of the poem composed by Gunnlaugr (Würth 1998, 81). The damage takes various forms. Chronologically first, and hypothetical rather than conclusively proven, are possible lacunae in the text of Merl that might have originated either memorially or scribally (see I 51 and II 48 Notes to [All]). Merl II is very much shorter, at 68 stanzas, than Merl I, at 103 – a disproportion that could have been caused by loss of text: for evidence of one quite extensive possible lacuna, extending to perhaps as many as fifteen stanzas, see Merl II 48 Note to [All]. On a smaller scale, Merl on several occasions lacks substantive material corresponding to individual clauses and sentences within the Prophecies. The variability of stanza length in Merl – normally eight lines but intermittently expanded to ten or twelve lines – opens the possibility that lines have been lost in transmission without leaving any detectable gaps in the Icelandic text, considered in isolation from DGB. Admittedly, as noted above, it can be difficult on occasion to distinguish between possible deliberate omission of material on Gunnlaugr’s part and loss in subsequent transmission. Another form of damage is specifically scribal. Although Haukr appears to have been a careful copyist, he commits evident errors from time to time, some corrected in his own hand but a significant number of others not detected (Hb 1892-6, xxxvi; see Notes to Merl I 21/5, I 35/5, I 36/2, I 38/5-6, I 44/1, I 79/7, I 81/7, I 84/4, Merl II 52/3, II 64/3, II 65/5, and II 66/7). Finally, massive damage has arisen thanks to conditions of preservation and curation of the codex (Jón Helgason 1960a, xxiv-xxv). The ink was already badly faded when Björn Jónsson of Skarðsá (1574-1655) gained access to the ms. During the time that Bishop Brynjólfur Sveinsson had the ms. on loan (approximately 1660-70), he had the ink refreshed where fading was compromising legibility. Refreshing on the folios containing Merl is said to have been done in part by Brynjólfur himself and in part by two legal officials (lögmenn), Sigurður Björnsson and Sigurður Jónsson (Handrit.is, accessed 05/07/2015). The incidence of refreshing is visible in the facsimile edition (Jón Helgason 1960a) and is thoroughly documented in the diplomatic transcription (Hb 1892-6, 272-83; cf. Skj A). Intermittent traces of the older ink beneath the refreshing indicate that not all the restoration was accurate (for instances see Merl II 6/1, II 14/3, II 16/1, 4, II 20/7, II 28/9, II 36/2, II 38/2, II 65/4); with access to new technology it is possible that further readings may be recovered, notably at Merl I 8/9-10. Thanks to some subsequent washing out of ink on the ms. pages, possibly in the course of attempted recovery of readings, a few words are now wholly illegible or barely legible, and at these points (indicated in the apparatus) Jón Sigurðsson’s edition (Bret 1848-9), supported by Hb 1892-6 and Skj A, becomes the primary, if not wholly reliable, witness to the text.
In the progress towards a modern edition the earliest known transcripts are due to Björn Jónsson. Although his autograph draft is lost, copies of copies of it survive (Jón Helgason 1960a, xxxii). He was unable to read any portions that are now illegible and his transcription, at least as preserved, contains obvious errors. Subsequent to Björn, and sometimes with the benefit of reference to these copies, several other attempts were made at partial transcriptions from Hb, but they do not add to our knowledge of the text (Jón Helgason 1960a, xxxv; Lavender 2006, 132-3) and are not cited in the present edition (for a partial listing of these mss see Merl 2012, 54-5). The decisive step toward an edition was made by Jón Sigurðsson, who transcribed and published virtually the entire poem and added the Latin text of DGB (using Giles 1844) for comparison. In the present edition, as noted, a few readings from Jón are cited selectively in places where Hb is now illegible, and designated ‘HbJS’. He incorporated in his edition some emendations contributed by Hallgrim (Hallgrímur) Scheving, schoolmaster at Bessastaðir, in the form of personal communications to Jón (cf. Jón Helgason 1960a, xxxvi). Scheving’s decisive contribution was to demonstrate that Gunnlaugr’s fidelity to DGB, sentence by sentence and sometimes word by word, could be of material assistance in attempts to repair damaged passages; for instances of emendations proposed by Scheving that deserve renewed consideration see Notes to Merl I 70/2, II 17/1-2 and II 31/8. Finnur Jónsson largely adopted Jón’s edition in Skj, with some new readings, emendations and interpretations. The posthumous edition of E. A. Kock (Skald) bases itself almost entirely on Skj, and it is clear from the mere scattering of notes in NN and FF and the lack of evident recourse to the ms., Bret or DGB that Kock never seriously engaged with this poem. The edition by Simone Horst (Merl 2012), with German translation, examines the ms. afresh and also considers select passages from DGB but fails to take proper account of a number of key matters: the metrics of the poem, the distinctive orthography of Hb, the contribution of Bret 1848-9 to the establishment of the text and current textual scholarship on DGB. In the present edition the two Parts are published in the order of their narrative chronology for the first time, a few further readings are recovered from Hb (see especially the Notes to I 72/6 and II 39/3), a number of new emendations and interpretations are proposed, the contribution of Bret is reassessed, systematic comparison with DGB is made stanza by stanza and DGB and Merl itself are explicated where necessary by reference to British historiography.
Merl has seldom been translated in other than excerpts. Aside from the editions, namely Bret 1848-9 and Skj B (into Danish), Merl 2012 (into German) and the present edition, there is only the translation into German by Stefanie Würth (1996, 98-120).