In the three stanzas edited below (Anon Gát), each line constitutes a ‘clue’ which, if understood, allows one to solve the riddle and reveal the name of a bird. Many involve word-play such as ofljóst ‘excessively clear’, in which a synonym of the clue is a homonym of the intended referent (e.g. st. 1/7 þjónustumey ‘serving-girl’ = þerna ‘serving girl’, ‘tern’); others rely on general knowledge including familiarity with the sagas and historical lore (e.g. st. 3/5 bana Gunnlaugs ‘Gunnlaugr’s slayer’ = Hrafn = hrafn ‘raven’; cf. Gunnlaugs saga ch. 12). The manuscript context of the gátur (‘riddles’), detailed below, suggests they were an intellectual pursuit among the learned. Some mss are annotated with solutions, not always in agreement with one another. There is no evidence for the existence of the gátur before c. 1500, the date of the earliest scribal hand, but an ofljóst riddle appears among the Heiðreks gátur (Gestumbl Heiðr 35VIII (Heiðr 82); see Note to st. 1/4 below), pointing to the existence of the form at least at an earlier date.
The oldest extant text appears in AM 625 4° (625), which is taken here as the main ms. The first part of the ms., which Kålund dates to c. 1300 (AÍ II, cxcv), contains, beside some shorter texts, Veraldar saga, Andrés saga postula, and Jóns saga baptista. The second part, dating from the fifteenth century (loc. cit.), contains various encyclopedic, computistic and liturgical texts, and the gátur. The gátur are not written by the scribe of the main text of this part of the ms; Kålund dates the hand of the gátur to c. 1500 (AÍ II, cxcvi n. 1).
AM 743 4°ˣ (743ˣ) is one of two main mss of the Y-version of the Laufás Edda. It was written by Ketill Jörundsson (1603-70), a priest at Hvammur in Dalasýsla, with some additions in other hands (see further LaufE 1979, 56-62). The gátur are in Ketill’s hand, but since they are only found in mss derived from this one, it is assumed they were not part of Magnús Ólafsson’s original Edda compilation (LaufE 1979, 56). Some of the gátur were annotated with their solutions by Árni Magnússon, who also added a fourth stanza (LaufE 1979, 58, 407); this is, however, incompatible with Norse poetry from the medieval period in terms of metre and diction, while l. 10 of st. 3 invites the addressee: gettu, hvat þeir heita ‘guess what they are called’, indicating the end of the series. The extra riddle is thus not included in this edition. Mss AM 738 4°ˣ and AM 755 4°ˣ contain a similar fourth post-medieval bird-riddle; AM 741 4°ˣ has this one plus a fifth attempt (LaufE 1979, 101, 122-4). These are again metrically irregular, but interesting for the evidence they provide for a continuing engagement with this sort of linguistic puzzle. Other LaufE Y-version mss which contain the riddles, or versions of them, are: AM 163 8°ˣ, Lbs 1199 4°ˣ, Lbs 756 4°ˣ, Lbs 818 4°ˣ, NM 64934, Lbs 1674 8°ˣ, ÍBR 35 4°ˣ, Lbs 1781 4°ˣ, Holm papp 67 folˣ, Lbs 636 4°ˣ and Lbs 1116 4°ˣ (LaufE 1979, 101-52). All the gátur were copied by Jón Ólafsson from Grunnavík in 1735 for his compilation of texts of SnE, now BL Egerton 642ˣ, and they appear in later copies and re-workings of this compilation in AM 436 folˣ, AM 974 4°ˣ, AM 429 folˣ (vol. 3), JS 199 4°ˣ, Lbs 217 4°ˣ and Lbs 3404 4°ˣ, none of which have any independent value (LaufE 1979, 141).
The third independent witness, AM 167 b 3 8°ˣ (167b 3ˣ), dates from the second half of the seventeenth century (Stories for all time database, accessed 22 July 2015). Folios 12-25 of this ms. were, prior to 1963, classed as fols 70-83 (the gátur on fols 72r-v) of AM 164 8°ˣ (ibid.) and are catalogued under that siglum in Kålund (1884-94, II, 424-5). Anon Gát 1-3 are preceded by the so-called Heiðreks gátur, Gestumbl HeiðrVIII, and followed by a post-medieval riddle on time (see further Love 2013, 233).
Lbs 1562 4°ˣ (1562ˣ; c. 1650-1799) is a ms. of the poems of the Poetic Edda, including some not found in the Codex Regius. Before these are Heiðreks gátur (Gestumbl HeiðrVIII), Anon Gát 1-3, plus the extra riddle modelled on them, and Anon SólVII. The ms. is of no independent value for the text of the gátur, and is not used in the present edition, although solutions to some of the riddles have been added in the scribal hand, and these are mentioned when of interest.
The gátur have been edited in Skj and Skald and are printed in SnE 1848 and LaufE 1979 from the text in 743ˣ. In the present edition, the text has been normalised to the standard of 1250-1300 in keeping with most of the editions in SkP VIII, including Gestumbl HeiðrVIII.