This interface will soon cease to be publicly available. Use the new interface instead. Click here to switch over now.

Cookies on our website

We use cookies on this website, mainly to provide a secure browsing experience but also to collect statistics on how the website is used. You can find out more about the cookies we set, the information we store and how we use it on the cookies page.

Data from Samnordisk runtextdatabas

login: password: stay logged in: help

Rǫgnvaldr jarl and Hallr Þórarinsson (RvHbreiðm)

12th century;

III. Háttalykill (Hl) - 84

No biography, so editor and volume are not set in this entry.

Háttalykill (‘The old key to verse-forms’) — RvHbreiðm HlIII

Kari Ellen Gade 2017, ‘ Rǫgnvaldr jarl and Hallr Þórarinsson, Háttalykill’ in Kari Ellen Gade and Edith Marold (eds), Poetry from Treatises on Poetics. Skaldic Poetry of the Scandinavian Middle Ages 3. Turnhout: Brepols, p. 1001. <> (accessed 5 July 2022)

stanzas:  1   2   3   4   5   6   7   8   9-10   11   12   13   14   15   16   17   18   19   20   21   22   23   24   25   26   27   28   29   30   31   32   33   34   35   36   37   38   39   40   41   42   43   44   45   46   47   48   49   50   51   52   53   54   55   56   57   58   59   60   61   62   63   64   65   66   67   68   69   70   71   72   73   74   75   76   77   78   79   80   81   82 

for reference only:  10x 

Skj: Rǫgnvaldr jarl og Hallr Þórarinsson: Háttalykill, o. 1145 (AI, 512-28, BI, 487-508); stanzas (if different): 1b | 2a | 2b | 3a | 3b | 4a | 4b | 5a | 5b | 6a | 6b | 7a | 7b | 8a | 8b | 9a | 9b | 10a | 10b | 11a | 11b | 12a | 12b | 13a | 13b | 14a | 14b | 15a | 15b | 16a | 16b | 17a | 17b | 18a | 18b | 19a | 19b | 20a | 20b | 21a | 21b | 22a | 22b | 23a | 23b | 24a | 24b | 25a | 25b | 26a | 26b | 27a | 27b | 28a | 28b | 29a | 29b | 30a | 30b | 31a | 31b | 32a | 32b | 33a | 33b | 34a | 34b | 35a | 35b | 36a | 36b | 37a | 37b | 38a | 38b | 39a | 39b | 40a | 40b | 41a | 41b

SkP info: III, 1001

notes: incomplete; in Orkn

old edition introduction edition manuscripts transcriptions concordance references search files


Háttalykill inn forni ‘The Old Key to Verse-forms’ (RvHbreiðm Hl) is a clavis metrica which in its present form consists of eighty-two stanzas. Metrically and thematically the stanzas can be divided into forty-one pairs, and each stanzaic pair displays one metrical peculiarity or verse-form and commemorates one legendary or semi-legendary hero or historical king. The poet’s identity cannot be established with absolute certainty, but it is very likely that there were two poets, namely, the Norwegian Rǫgnvaldr jarl Kali Kolsson of Orkney (Rv) and the Icelander Hallr Þórarinsson breiðmaga ‘(Son) of Broad-belly’ (Hbreiðm). According to Orkneyinga saga, the two composed a poem together, which was called Háttalykill inn forni (Orkn ch. 81, ÍF 34, 185): … var hann lengi síðan með Rǫgnvaldi jarli. Þeir ortu báðir saman Háttalykil inn forna ok létu vera fimm vísur með hverjum hætti. En þá þótti of langt kveðit, ok eru nú tvær kveðnar með hverjum hætti ‘… he [Hallr] stayed for a long time afterwards with Rǫgnvaldr jarl. Together they composed Háttalykill inn forni and they had five stanzas for each verse-form. But then it seemed too long a composition, and now there are two stanzas for each verse-form’. Chapter 81 of Orkn is transmitted in two mss, Flateyjarbók (Flat) and UppsUB R 702 4ˣ (R702ˣ), as well as in an early Danish translation (Holm papp 39 folˣ), and the texts of these versions differ somewhat in terms of the title of the poem (hꜳttalykil hinn forna Flat(138v); hatta tal hid forna R702ˣ(42v)) and the number of stanzas illustrating each metre in the final version (tuær ‘two’ Flat; iij R702ˣ; Holm papp 39 folˣ has tre ‘three’ as well; see ÍF 34, 185 n. 2). The title Háttatal it forna ‘the old enumeration of verse-forms’ in R702ˣ can most likely be explained by the fact that Magnús Ólafsson, the scribe of that ms., was familiar with Snorri Sturluson’s Háttatal (SnSt Ht) and he must have changed the name of the poem accordingly (see Hl 1941, 5-6; ÍF 34, 185 n. 1). The number of stanzas for each metre given in the mss (two or three) is more problematic. All versions agree that the poem originally had five stanzas illustrating each verse-form and that the poets decided to shorten it because of its length. The paradosis of the three mss shows that Flat is lower on the stemma than R702ˣ and Holm papp 39 folˣ, however, and moreover, the latter two mss belong to different branches of the stemma (see the stemma in ÍF 34, cxxvi; SkP II, lxxv-lxxvi). Hence it looks as though the numeral ‘two’ is a mistake that originated in Flat, and that the number of stanzas per metre in the revised version of Hl was three and not two as in its present version (see the discussion in Hl 1941, 6-7, 118, 140). We cannot therefore know for certain whether the extant Hl is the poem composed by Rǫgnvaldr and Hallr mentioned in Orkn, but detailed analyses of palaeographic and dialectal features lend credence to such an attribution (see Hl 1941, 99-118, 140-2 and the discussion below).

If Hl was indeed the creation of Rǫgnvaldr and Hallr, the terminus ante quem for the composition of the poem is 1158/59 (Rǫgnvaldr died on 20 August 1158 or 1159; see his Biography in SkP II). Hallr arrived in Orkney prior to Rǫgnvaldr leaving for Norway in 1148, and according to Orkn (ÍF 34, 185), he stayed with Rǫgnvaldr for a long time after his arrival. Finnur Jónsson (LH 1894-1901, II, 35) dates Hallr’s Orcadian sojourn to 1140-8, and in Skj A he gives 1145 as an approximate date for the poem. Other scholars have favoured a date of composition after Rǫgnvaldr’s return to Orkney from his crusade in 1153 (e.g. de Vries 1938, 720, 733 n. 73; Kuhn 1983, 317).

The poem is transmitted without prose in two mss, Holm papp 25 8°ˣ (papp25ˣ) and UppsUB R 683ˣ (R683ˣ), both dating from c. 1665 and in the hand of the Icelander Jón Rugman. The superior ms., papp25ˣ (the main ms. for the present edition), was not discovered until 1937, when Jón Helgason found it in the Royal Library in Stockholm (for a description of that ms. and its content, see Jón Helgason in Hl 1941, 7-16). Hence editions of Hl that predate 1937, including Skj and Skald, are based on the inferior ms. R683ˣ, which greatly diminishes their value for later editions of the poem (see below and the discussions in the Notes to the individual stanzas). In papp25ˣ Hl is recorded on leaves 33r-43r, and the heading (at 33r) is Fragmenta af hatta lijkle ‘Fragments of háttalykill’. It is preceded by excerpts from Snorra Edda and from Arngrímur Jónsson’s Crymogæa (in the hand of Jón Rugman) and by some poetic stanzas in other hands. At the end of the poem (43r), Rugman continues with a citation from Ole Worm (1650, 181; derived from a treatise sent to Worm by Magnús Ólafsson, the scribe of ms. R702ˣ of Orkn; see Hl 1941, 6, 13), which credits Rǫgnvaldr jarl (Rogvaldus quidem Orcadum comes) with the composition of claves metricae (clavem Rhythmicus [sic]; clearly stemming from Magnús’s own familiarity with the text of Orkn). At the end of the citation in papp25ˣ(43r) Rugman adds forte Hanc? ‘perhaps this one?’. As Jón Helgason (Hl 1941, 13) points out, Rugman’s exemplar could not have contained the name of the Hl poet (although the title of the poem must have been given there), and he speculated about his identity. After the excerpt from Worm (1650, 181) Rugman continues (at 43v) with a citation from Arngrímur’s Crymogæa unrelated to Hl, and on leaves 44r-v he gives somewhat confused information about the exemplar he used for his copy of Hl (printed in Hl 1941, 13-14, and discussed in detail by Jón Helgason, ibid.). The main points of Jón Helgason’s discussion can be summarized as follows (Hl 1941, 14-15): Rugman copied an old and partly illegible ms. (exscriptum), which contained corrections (by a corrector) that Rugman believed to have come from an even older ms. (prototypus). Rugman first made a quick copy of the text, and later he went back to look more carefully at the exemplar to see whether he could read more. He tried to decipher what was legible and marked illegible places and letters by underscorings, dots and short dashes. According to Jón Helgason (Hl 1941, 12), Rugman must have copied the ms. during his stay in Copenhagen in 1665. The old ms. that was his exemplar most likely came from a private collection in Copenhagen, perhaps the collection of Villum Worm, and it could have been destroyed in the fire in Copenhagen in 1728.

In an attempt to arrive at a date and provenance for Rugman’s exemplar, Anne Holtsmark (Hl 1941, 99-115) gives a detailed analysis of the orthography and language of Rugman’s copy of Hl in papp25ˣ (see also Jón Helgason in Hl 1941, 15-16). The difficulty of that undertaking was compounded by the fact that Rugman was not trained in palaeography and, furthermore, his knowledge of Old Norse left a lot to be desired. Hence his transcriptions contain quite a few mistakes and also many later Icelandic forms and spellings. Nonetheless, Holtsmark (Hl 1941, 114-15) is able to conclude that Rugman’s exemplar must have been a very old ms. (c. 1180-90) probably written in south-eastern Norway, although she does not exclude the possibility that it could come from Orkney.

Rugman made a second copy of Hl in ms. R683ˣ(125r-134v), which was the basis for all editions predating 1937 (for a discussion of R683ˣ, see Hl 1941, 16-19). In that ms. the Old Norse text is written on the left-hand side of each leaf, and on the right side Rugman provided his own Latin translations of the stanzas. As Jón Helgason (Hl 1941, 17) points out, the Latin translations are quite worthless and Rugman often had problems understanding the Old Norse text. He also omitted translations of stanzas that were damaged in his transcriptions. In R683ˣ the poem is introduced by the heading Hatta=lijkill Ragnvaldar jarls (rendered in Latin as Clavis Rhÿthmica Comitis Rognvaldi). The attribution to Rǫgnvaldr is no doubt due to Rugman’s own speculations about the identity of the Hl poet and based on Magnús Ólafsson’s reference to Rǫgnvaldr in Worm 1650, 181 (see above). Jón Helgason’s comparison of the texts in papp25ˣ and R683ˣ (Hl 1941, 19-21) shows that R683ˣ is a copy of papp25ˣ and that it is clearly the inferior ms. There are textual divergences between the two mss, and Rugman added quite a few conjectural corrections to the original text in R683ˣ, many of which are erroneous, and the ms. contains more late Icelandic forms than papp25ˣ. According to Jón Helgason (Hl 1941, 20-1), the plethora of conjectural readings in R683ˣ reflects Rugman’s attempt at creating a ‘clean’ copy – sometimes he hit upon the correct forms, but more often than not he failed to do so. It is not clear whether Rugman still had access to his old exemplar when he made the R683ˣ copy of the text in papp25ˣ. While there is no reason to conclude with certainty that the exemplar was still in his possession, in a few instances R683ˣ contains readings superior to those of papp25ˣ. That circumstance led Jón Helgason (Hl 1941, 21) to suspect that Rugman may have made the R683ˣ copy prior to leaving Copenhagen in 1665, and that he could have had access to the exemplar and was able to make out some words and letters that he could not read earlier.

As noted above, in the extant version of Hl each verse-form is illustrated by two stanzas and in both mss each set of stanzas is assigned a number. In earlier editions of the poem, the stanzas were conventionally numbered sts 1a, 1b, 2a, 2b, etc. That format proved to be incompatible with an earlier layout of the database of the skaldic project, and the decision was made to assign a separate number to each stanza (sts 1, 2, 3, etc.). Hence, in Skj, Skald and Hl 1941, the numbers run from sts 1a-41b, whereas the present edition has sts 1-82. In the discussions of the order of the stanzas below the old numbers are given in parentheses for the sake of convenience.

The order of stanzas in papp25ˣ is quite chaotic and resulted from Rugman’s inability to decipher many places in the text of his exemplar. His transcription begins with st. 1 (1a) and he then skips to sts 29 (15a)-68 (34b). After st. 68 (34b) he records sts 73 (37a)-82 (41b) before he proceeds to copy stanzas which are not preceded by the name of a metre (Horum nomina desunt ‘The names of these are missing’), that is, sts 69 (35a)-72 (36b). After st. 72 he gives the remaining stanzas in the following order: 7-8 (4ab), 11-12 (6ab), 17 (9a)-20 (10b), 23-24 (12ab), 2 (1b)-6 (3b), 9-10 (5ab), 13 (7a)-16 (8b), 21-22 (11ab), 25 (13a)-28 (14b). Because each set of stanzas is clearly numbered, this arrangement does not present any problems. In R683ˣ the stanzas are given in their numerical order except that sts 67-8 (29ab) are placed between st. 52 (26b) and st. 53 (27a), possibly because Rugman attempted to restore the genealogical sequence of champions commemorated in the poem (so Jón Helgason, Hl 1941, 20).

As stated above, each pair of stanzas in Hl commemorates legendary heroes or historical kings. The legendary heroes are known from eddic poetry, the fornaldarsögur, Skjǫldunga saga and Saxo’s Gesta Danorum, and their exploits are recounted in sts 1-58. The heroes commemorated are the following: Sigurðr Fáfnisbani ‘Slayer of Fáfnir’ Sigmundarson (sts 3-4); Hǫgni Gjúkason (sts 5-6); Gunnarr Gjúkason (sts 7-8); Helgi Hundingsbani ‘Slayer of Hundingr’ Sigmundarson (sts 9-10); Ragnarr loðbrók ‘Shaggy-breeches’ (sts 11-12); Ælla of Northumbria (sts 13-14); Ragnarr loðbrók’s sons Ívarr inn beinlausi ‘the Boneless’ (sts 15-16), Bjǫrn járnsíða ‘Ironside’ (sts 17-18), Sigurðr ormr-í-auga ‘Snake-in-the-Eyes’ (sts 19-20) and Hvítserkr (sts 21-2); Svipdagr (sts 23-4); Hjalti (?) (sts 25-6); Haki (?) (sts 27-8); Hagbarðr (sts 29-30); Friðleifr Fróðason (sts 31-2) and his sons Áli inn frœkni ‘the Bold’ (sts 33-4) and Fróði (sts 35-6); Helgi Hálfdanarson Fróðasonar (sts 37-8); Haraldr hilditǫnn ‘War-tooth’ (sts 39-40); Hringr Randvésson (?) (sts 41-2); Angantýr (sts 43-4); Heðinn, Hǫgni and Hildr (sts 45-6); Hrólfr kraki ‘Pole-ladder’ (sts 47-8); Hjálmarr inn hugumstóri ‘the Great-minded’ (sts 49-50); Eiríkr inn sigrsæli ‘the Victorious’ Bjarnarson (sts 51-2); a sea-king Óláfr (?) (sts 53-4); Gautrekr (sts 55-6); Styrbjǫrn inn sterki ‘the Strong’ Óláfsson (sts 57-8). With st. 59 the poem turns to the kings of Norway (and Denmark), beginning with Haraldr hárfagri ‘Fair-hair’ Hálfdanarson (sts 59-60) and proceeding in (almost) chronological order on to Eiríkr blóðøx ‘Blood-axe’ Haraldsson (sts 61-2), Hákon inn góði ‘the Good’ Haraldsson (sts 63-4), Haraldr gráfeldr ‘Grey-cloak’ Eiríksson (sts 65-6), Sveinn tjúguskegg ‘Fork-beard’ Haraldsson (sts 67-8), Knútr inn ríki (Cnut the Great) Sveinsson (sts 69-70), Óláfr Tryggvason (sts 71-2), Óláfr Haraldsson (sts 71-2), Magnús inn góði ‘the Good’ Óláfsson (sts 73-4), Haraldr harðráði ‘Hard-rule’ Sigurðarson (sts 75-6), Óláfr kyrri ‘the Quiet’ Haraldsson (sts 79-80) and Magnús berfœttr ‘Barelegs’ Óláfsson (sts 81-2). In terms of content the stanzas are highly stereotyped and tend to focus on the warlike exploits of the champions they commemorate. It is clear that the poet or poets knew the names of these heroes and kings, but very little factual information is otherwise provided about them. The names of epic battles are omitted (with the exception of the battle of Hafrsfjorden (sts 59-60) and, indirectly, the battle of Svolder (sts 73-4)), as are place names and, with a few exceptions, patronymics. It looks as though the focus was on form rather than on content. A total of twelve stanzas (sts 11-22) are devoted to characters from the stories and poetry about Ragnarr loðbrók, a hero who is also commemorated in the runic inscriptions in Maeshowe, Orkney, some of which could have been carved by Rǫgnvaldr and his crusaders (see Note to Rv Lv 1/3 and Barnes 1994, 40-1). De Vries (1938, 720-1) suggests that Rǫgnvaldr must have been very familiar with the stories about Ragnarr and his sons, which could account for the high proportion of poetry accorded to them in Hl (for a comparison between the language of the Maeshowe inscriptions and the language of ms. papp25ˣ, see Hl 1941, 115-19).

The poet(s) borrowed from earlier skaldic poems (see de Vries 1938, 714-15) and in some instances there is an almost verbatim lifting of poetic lines (see Hl 1941, 135-6 and Notes to the individual stanzas), but, as observed by Holtsmark (Hl 1941, 136-9), the most peculiar feature of the poem is that, by repeating words, phrases and whole lines, it borrows most blatantly from itself (see Notes to the individual stanzas).

Each pair of stanzas in Hl illustrates a specific metre or metrical feature, some of which are not previously attested in Norse territory and not found later outside of the claves metricae (for a discussion of Hl and Snorri Sturluson’s Ht (SnSt Ht) see Introduction to that poem). It is quite likely that some of the metrical innovations are based on foreign models (see de Vries 1938, 21-4;  Hl 1941, 121-34), but in other instances the metrical (and syntactic) variations represent a systematisation of metrical features found in the poetry of earlier skalds (see Notes to the individual stanzas). In that respect the poem is the forerunner to Snorri’s Ht. Unlike Ht, however, the different verse-forms are not introduced in any systematic manner. Aside from the names of the metres preceding many of the stanzas, there is no prose commentary accompanying the poetic texts in Rugman’s two mss, and it is clear that his transcriptions represent his attempt to reproduce what was in his exemplar. Rugman provides a total of thirty-five names of metres, of which twenty-nine also occur in Ht. When his exemplar did not contain the name of a verse-form, Rugman added the rubrics Titulus deest (papp25ˣ) or Inscriptio deest ‘The title is missing’ (R683ˣ). The table below provides an overview of the correspondences between the metres and the terminology used in the two metrical treatises. The left-hand column contains the terms used in the numbered stanzas of Hl and the corresponding stanzas in Ht are given in the right-hand column. Square brackets mean that the metre is attested in both, but that the name is missing or fails to correspond to the name of the metre in the other treatise.

1-2. ljóðsháttr ‘song’s form’ 100
3-4. kviðuháttr ‘poem’s form’ [102]
5-6. [No title] (= dróttkvætt ‘court metre’) 1-8
7-8. draughent ‘ghost/trunk-rhymed’ 65
9-10. [No title] (= stýfðr ‘apocopated’) 51
11-12. [þríhent ‘triple-rhymed’] (name added from Ht) 36
13-14. belgdrǫgur ‘bellows-drawings’ [82]
15-16. munnvǫrp ‘mouth-throwings’ 66
17-18. inn dýri háttr ‘the ornate form’ 37
19-20. inn grœnlenzki háttr ‘the verse-form from Greenland’ 71
21-2. [No title] (= runhent ‘end-rhymed’) [88]
23-4. [No title] [48, 74-5, 77]
25-6. tøgdrápu háttr ‘journey-poem form’ 68-9
27-8. kimblabǫnd ‘bundle-bonds’ 61
29-30. háhent ‘high-rhymed’ (= náhent ‘close-rhymed’) [75]
31-2. hrynhent ‘flowing-rhymed’ 62-4
33-4. rekit ‘extended’ [90]
35-6. detthent ‘stumbling- or falling-rhymed’ 29
37-8. bálkarlag ‘section’s metre’ 97
39-40. refrún in minni ‘the lesser fox-secret’ (= refhvǫrf ‘fox-turns’) [22]
41-2. sextánmælt ‘sixteen-times spoken’ 9
43-4. hnúfu háttr (?)
45-6. greppaminni ‘poets’ reminder’ 40
47-8. runhent ‘end-rhymed’ [87]
49-50. hálfhnept ‘half-curtailed’ 77
51-2. háttlaus ‘formless’ 67
53-4. Haddarlag (?) (= Haðarlag ‘Hǫðr’s metre’) 79
55-6. refrún in meiri ‘the greater fox-secret’ [20-1]
57-8. iðurmælt ‘repeatedly spoken’ 47
59-60. langlokum ‘with late closures’ 14
61-2. alstýft ‘completely apocopated’ 50
63-4. flagðalag ‘ogresses’ metre’ 34
65-6. dunhent ‘echoing-rhymed’ 24
67-8. tilsegjandi ‘annotating’ (= tilsagt ‘annotated’) 25
69-70. [No title] [78, 17-18]
71-2. [No title] (= draugsháttr ‘ghost’s verse-form’) 30
73-4. konungslag ‘king’s metre’ [63]
75-6. áttmælt ‘eight-times spoken’ 10
77-8. hálfhnept ‘half-curtailed’ (= alhnept ‘completely curtailed) [78]
79-80. álagsháttr ‘extension’s form’ 27
81-2. skjálfhent ‘tremble-rhymed’ 28, 35

As this table shows, most of the metres in Hl (except hnúfu háttr, sts 43-4) are also attested in Ht, and there is a fairly strong correspondence as to the metrical terminology employed in the two treatises as well. The terms belgdrǫgur ‘bellows-drawings’ (sts 13-14) and konungslag ‘king’s metre’ (sts 73-4) are not used in Ht, but the metres correspond to Ht 82 and 63, respectively. The metre refrún ‘fox-secret’ (sts 39-40, 55-6) appears in Ht as refhvǫrf ‘fox-turns’ (Ht 17-22), and kviðuháttr ‘poem’s form’ (sts 3-4) is attested in Ht 102, although Snorri assigns no name to that metre. Sometimes Rugman appears to have misread the name of a metre (or it was garbled in his exemplar): sts 29-30 háhent ‘high-rhymed’ corresponds to Ht 75 náhent ‘close-rhymed’ and sts 77-8 hálfhnept ‘half-curtailed’ is likely to be a misreading of alhnept ‘completely curtailed’ (Ht 78). In one instance (sts 11-12), the metre does not have a name in papp25ˣ (Titulus deest, rendered in R683ˣ as Inscriptio deest), and Rugman added the caption þríhent ‘triple-rhymed’ in R683ˣ, most likely from the heading in ms. U of SnE (Ht 36).

The poem is incomplete in both mss and it ends at st. 82/2, after which Rugman adds Plura desiderantur ‘More are lost’ (papp25ˣ) or Multa desunt ‘Many are missing’ (R683ˣ). It is impossible to know how many stanzas are missing at the end, but if the list of Norwegian kings formed the last part of the poem, and the extant version of Hl is indeed the composition of Rǫgnvaldr and Hallr, it could not have extended past 1158/59. The last king to be commemorated (sts 81-2) is Magnús berfœttr Óláfsson (d. 1103), and he was succeeded on the Norwegian throne by his sons Ólafr (d. 1115), Eysteinn (d. 1123) and Sigurðr jórsalafari ‘Jerusalem-farer’ (d. 1130). Subsequent Norwegian kings include Haraldr gillikristr ‘Servant of Christ’ Magnússon (d. 1136), Magnús blindi ‘the Blind’ Sigurðarson (d. 1139) and Sigurðr slembidjákn ‘Fortuitous-deacon’ (?) Magnússon (d. 1139), as well as the sons of Haraldr gilli, Sigurðr (d. 1155), Eysteinn (d. 1157) and Ingi (d. 1161). Holtsmark (Hl 1941, 119) suggests that the poem could have ended as it started, namely by commemorating the deeds of a ‘Sigurðr’ (sts 3-4). If that were the case, the best candidate would be Sigurðr jórsalafari, the crusader king and royal benefactor who gave Rǫgnvaldr dominion over half of Orkney (Orkn ch. 61, ÍF 34, 140-1).

Hl is the earliest clavis metrica to be composed in Norse territory and, as Holtsmark (Hl 1941, 121-2) notes, the title of the poem, Háttalykill ‘key to verse-forms’ (= clavis metrica, clavis rhytmica), suggests that this poetic exercise may have been prompted by foreign, learned influence (for an overview of comparable early continental poems, see Hl 1941, 121-4). Holtsmark (Hl 1941, 118-34) goes on to discuss in detail possible foreign models for the various Old Norse metres exemplified in Hl. In the present edition, the results of her investigation are addressed in the Notes to each stanza when relevant. In one of his lausavísur (Rv Lv 1/4, 8II), Rǫgnvaldr boasts that bók ‘the book’ and bragþættir ‘poems’ are among the skills that he is proud to master, and there can be no doubt that such metres as refrún ‘fox-secret’ (sts 39-40), greppaminni ‘poets’ reminder’ (sts 55-6) and, in particular, hnúfu háttr (?) (sts 43-4) may indeed have been prompted by foreign influence. It is equally clear, however, that most metres in Hl are variants in which single metrical peculiarities found in indigenous Norse poetry have been stylised and used systematically throughout entire stanzas by one or more poets who were in complete command of the intricacies of native poetry. This circumstance raises the question that was posed at the beginning of this Introduction, namely, whether the poem was the composition of one or two poets. In her detailed discussion of this issue, Holtsmark (Hl 1941, 135-42) concludes that the latter indeed appears to be the case. The pairs of stanzas illustrating each verse-form display stylistic variation, which indicates that they are not likely to have been composed by the same poet and, furthermore, the internal borrowing of words, phrases and whole lines that characterises the poem is suspect. To that we may add that the first poet sometimes seems to issue a challenge to which the second poet responds (see e.g. sts 55-6 and Notes), and in one instance the second poet apparently corrects the first poet when he disagrees with his presentation of the king commemorated in his stanza (see sts 79-80). Holtsmark (Hl 1941, 139-40) believes that Hl may have originated as a poetic competition between two poets, and that suggestion is very attractive. Not only are such competitions well known from the royal compendium Morkinskinna (see e.g. Hharð Lv 5, 10-11II, ÞjóðA Lv 3, 4II, Þfisk Lv 1-3II, Anon (HSig) 3-4II), but they also appear to have been a pastime of Rǫgnvaldr and his poets (Orkn ch. 85, ÍF 34, 202): Þat var einn dag um jólin, at menn hugðu at tjǫldum. Þá mælti jarl við Odda inn litla: ‘Gerðu vísu um athǫfn þess manns, er þar er á tjaldinu, og haf eigi síðarr lokit þinni vísu en ek minni. Haf þú ok engi þau orð í þinni vísu, er ek hefi í minni vísu’ ‘It happened on a day during Christmas that people were looking at the tapestries. Then the jarl said to Oddi inn litli: “Compose a stanza about the behaviour of that man who is depicted on the tapestry, and do not take more time to complete your stanza than I take to complete mine. Also, do not use any of those words in your stanza which I have in my stanza”’. For the ensuing stanzas, see Rv Lv 13II and Oddi Lv 1II (see also the stanzas composed about Ermingerðr of Narbonne, Rv Lv 16-17II, Oddi Lv 3II, Árm Lv 3II). One can readily imagine that a poem such as Hl could have been the result of a similar poetic competition in which the first poet challenges the second to compose in the same verse-form and about the same hero or king as he does, and, in the case of Hl, ‘use the same words in your stanza as I have in my stanza’. There is nothing in the language of the stanzaic pairs to suggest which of the two poets (if they were the Norwegian Rǫgnvaldr and the Icelandic Hallr) may have been the a-poet and which the b-poet (see Hl 1941, 141). If the original composition consisted of odd-numbered sets of stanzas rather stanzaic pairs, however, it could well be that the competition was conceived of as a type of relay, in which the second and last poet was the first to pick up the baton and start off the next set of stanzas.

Jón Sigurðsson was the first to produce a copy of R683ˣ (JS 404 4°ˣ; written in the summer of 1841, see Hl 1941, 22). Jón’s copy, which was partly normalised and contained some errors, was lent to Sveinbjörn Egilsson, who used it as the basis for the text of Hl in his 1848 edition of Snorra Edda (SnE 1848, 239-48, 252). That edition, which is only partial and omits a number of stanzas that were damaged in Rugman’s transcriptions, has been consulted for the present edition but it is only used selectively. Finnur Jónsson’s (Skj) and Kock’s (Skald) editions of the poem are also based solely on the inferior ms. R683ˣ. Because neither editor had access to papp25ˣ, their interpretations are more often than not based on Rugman’s conjectural readings and have not been discussed in extenso in Notes to the present edition. Hl 1941 presents a diplomatic text of papp25ˣ (plus an appended facsimile of that ms.) with accompanying Notes to each stanza. The present edition is greatly indebted to the work of Jón Helgason and Anne Holtsmark. The only drawback with the 1941 edition is that it contains a number of highly conjectural reconstructions of stanzas that are incomplete in Rugman’s copies. Unless these reconstructions have support in the mss or can be ensured by metre, they have not been considered in the present edition.

The text in the main ms. papp25ˣ (supplemented by variants from R683ˣ) presents quite a challenge in terms of normalisation. Rugman’s exemplar was a very old, probably Norwegian ms. with Norwegian orthography, and his transcriptions contain quite a few mistakes. In addition, he often introduced later Icelandic forms in his texts. The language of the present edition has been normalised to pre-1200 Icelandic, and the decision was made to normalise ex silentio spellings that could in any way be explained as dialectal or orthographic (Norwegian) or later Icelandic and to list the ms. forms in the Readings (with accompanying discussion in the Notes when necessary). Hence the Readings section contains more diplomatic readings than is usual in SkP editions. In some instances it is difficult to determine whether a particular form is due to orthographic/dialectal variation or to scribal error (e.g. loss of final ‑r, graphemic hypercorrection or simplification of a final consonant cluster). If the resulting form is grammatically ambiguous (e.g. final n. nom. sg. <tt> rendered as <t>), the reintroduction of the correct grammatical form has been treated as an emendation and marked as such in the Text and in the Readings. Rugman also marked places that were illegible to him by uneven sequences of dots or short dashes. In the present edition, those sequences are rendered uniformly by three full stops (‘…’) in the Text, Prose Order and Translation.

Runic data from Samnordisk runtextdatabas, Uppsala universitet, unless otherwise stated