This book's blurb states that a new edition of the twelfth-century Old Norse-Icelandic skaldic poem Geisli has been "long needed by scholars". There are several reasons why this is so. Geisli, a stately 71-stanza drápa (an encomium in which subsections are separated by refrains) on the miracles of St Óláfr by the Icelandic skald Einarr Skúlason, is the earliest drápa transmitted as a whole, and one of the few about whose performance circumstances much is known. It is also an early and influential example of a skaldic poem with Christian subject-matter, a subgenre which was to dominate the later centuries of skaldic production. A glossed and annotated English language edition with commentary, of which Chase's is the first, will not only bring this poem to a wider audience of researchers outside the specialised field of skaldic studies, but should also be pedagogically valuable. Much of this book - especially the introductory essay, packed with intriguing insights - fulfils this promise. However, its usefulness is compromised both by methodological weaknesses and by errors and inconsistencies stemming from Chase's (and his editors') failure to harmonise multiple stages of revision of the text.
The book opens with a brief "Introduction", comprising information about manuscripts, previous editions, author, date, sources, metrics and diction. The wayGeisli blends skaldic traditions with Christian learning - forms, figural schemes and subject-matter - is a key focus of the edition. This topic is touched on here in the comparison of Geisli's drápa form with "the office of matins for a saint's day, in which the lectiones are separated and framed by an antiphon" (15) - though as refrains seem to have been traditional in all kinds of drápur, this liturgical echo may not have been so striking for the poem's hearers - and perhaps more convincingly, of Einarr's chronological arrangement of Óláfr's miracles with the similarly-organised lists appended to saints' lives (17).
A longer "Appreciation" expands upon the learned Christian background to Geisli in the course of a stanza-by-stanza commentary on its structure and subject-matter. The miracle stories which make up the bulk of the poem are explained: essential information if the reader is to make sense of Einarr's often allusive references to them. Chase makes many interesting and persuasive points about the theology of the poem, for example Einarr's use of the sun/sunbeam figure to persuade his listeners of the "mutual indwelling" (23) of God and Óláfr, and the role of the skald's allusions to the Gospel of John. Theological themes are taken up again in the Commentary, especially on sts. 1-6 (though see also the notes on 11.2, 16.7-8, 28.3, 63.8 and 65). Chase describes possible sources and analogues for Einarr's miracle stories; usually written texts, though in one case (the miracle of Óláfr's sword Hneitir) Einarr seems to have been the one who introduced the story into the written tradition (41-2). Altar frontals are mentioned several times in the course of a discussion of the typological links between Christ and Óláfr, and it would have been helpful had space been found for images, as the discussion of the St Olav altar frontal in Trondheim cathedral in particular (34) is slightly difficult to follow as it stands.
Three pages of "Headnotes to the Text" (45-7) describe the orthography of the base manuscript transcription (though not of the normalised text) and explain the parts of the textual apparatus: "textual notes", "lexical variants", "phonological and spelling variants" and "editorial variants", i.e. readings from previous editions which differ from Chase's. One cannot help thinking that the editor has made a rod for his own back by dividing the manuscript variants into three categories, and the apportioning of readings between them is not entirely consistent.
The edited text follows. For each of Geisli's 71 stanzas a transcription of the Flateyjarbók manuscript's text, laid out as eight lines of verse with editorial interventions between angled brackets, is presented side by side with a normalised ON text. A reordering of the stanza's words into prose syntax and an English translation of this come immediately below, and below them the textual apparatus. This layout is clear and user-friendly, though its execution is sometimes marred by errors. The edited stanzas are followed by a "Commentary", focusing on textual cruces, alternative manuscript and editorial readings, lexical borrowings, and explication of individual kennings. Chase is inclined to view many of Geisli's kennings as Gelegenheitskenningar "occasional kennings", i.e. as having meanings which are determined by their particular contexts. This often produces interesting interpretations (the observation that "Traditional kennings for the retainer who enjoys the material generosity of his lord are used in a new context to represent those who are granted favours by God through Óláfr's intercession", 19) though sometimes the argument seems overstretched (e.g. the comments on 32.1, 2, 4, p. 148 and 20.6-7, p. 142). But surely the "perplexing" (142), "strange" (143), or "jarring" (144) relation of some of Einarr's kennings to their context is a manifestation of the necessary conventionality of the kenning system? If every kenning was a Gelegenheitskenning the system wouldn't work (cf. the recent article by Gary Holland in ANF 120 (2005) arguing persuasively for kennings as "semantic formulae"). A glossary, a list of abbreviations of secondary works, notes to the introductory sections, a bibliography and an index of proper names, technical terms, kennings and manuscripts round off the book.
This new edition is thus a substantial work. Given this, it is disconcerting how little space Chase devotes to justifying his methodological decisions. His edition seems conceived as a re-assessment of the methodology embodied in the standard edition of the skaldic corpus, the four volumes of Finnur Jónsson's Skj (Den norsk-islandske skjaldedigtning, AI-II, BI-II, 1912-15). Finnur's B-text, or critical edition, as is well known, is highly eclectic. It also contains many readings which lack any manuscript support, instead having their origins in Finnur's ingenuity or that of his predecessors. Neither of these kinds of editorial intervention are marked in the Skj B-text and it is up to the reader to compare this text with the diplomatic transcription in Skj A, which has often been criticised for its unreliability. Skj A takes F (Flateyjarbók) as its base text for Geisli, and Chase points out some errors in its transcriptions from this manuscript on p. 8 in the course of his survey of previous editions. They seem, to be fair, a scanty crop over 71 stanzas. Nor are Skj A's misplaced accents, however irritating, very significant, as the marking of vowel length in medieval Icelandic manuscripts is so unreliable.
Chase does not discuss Skj B in his survey of previous editions. This is a surprising omission as for all its faults it is the standard edition, and in the context of skaldic studies Chase's avoidance not only of editorial surmise lacking manuscript support, but also of introducing readings from other manuscripts into his edited text, strikes the reader as a reaction against Skj's methodology. Chase's edition is instead "a corrected base text, with variant readings from the other manuscripts" (4). This is an entirely reasonable editorial methodology in itself, but its pitfalls soon become apparent where this particular poem is concerned. As the editor observes, neither of the two medieval mss which contain all or almost all of the poem, Bergsbók and Flateyjarbók, has a good text. His justification for choosing Flateyjarbók (F) over Bergsbók (B) runs as follows: "three stanzas are missing from F, but the arrangement of stanzas in that manuscript is better than in B. Both manuscripts are highly corrupt, but the language and orthography of F are more regular, and for these reasons I have chosen it over B as the base manuscript" (4). Regularity of spelling seems a curious criterion for the choice of the base ms. for a critical edition, but it is perhaps explained by Chase's desire to prioritise the non-normalised form of the text. The intention seems to be that there should be as few obstacles as possible between the reader and the late medieval realisation of the poem as preserved in F. The normalised text is described as merely an aid to looking up words in dictionaries (4), and is punctuated very lightly, also with the intention of not constraining the reader's interpretative activity (5).
The fundamental problem is that F's text of the poem often seems markedly inferior to that in B. On many occasions (I counted 28) Chase admits in his Commentary that the F reading presented in his main text is less good than the one in B. Sometimes following F's readings necessitates implausibly convoluted syntax (e.g. 22b, 40, 52b, 66a; the justification of st. 40's complexity on p. 151 is weak), though the minimal punctuation in the normalised verse text means this is less immediately apparent than it ought to be. In other cases it means the edition reproduces F's metrical faults, typical of fourteenth century scribal renderings of skaldic verse, such as hypermetric lines (often due to the editor's avoidance of vowel elision, or bragarmál, but sometimes a matter of extra syllables which cannot be metrically resolved) and missing or misplaced rhymes (e.g. 8.6, 11.8, 62.7, 65.7, 71.5). F's text in 14.3 (pl. þegnar as subject of sg. muna "will not") and 45.1 (sánom when the prose syntax and sense of the half-stanza require B's þann acc.) is grammatically untenable and must be rejected. And often F is simply clumsier than B: see for example 11b, where F requires us to understand a noun to go with íþessum (Chase translates as "among these [people]"), while B's íþessu (sg.) modifies ríki 9realm9; 57.7, where F has the otherwise-unattested adjective haldframr "? durably outstanding" in place of B's commonplace hjaldrframr "outstanding in battle"; 71b, where F's segir "say"and elskik "I may love" are surely scribal simplifications of B's imperatives, unexpected in the encomiastic context, segið and elskið (F's scribe rarely if ever cliticises personal pronouns elsewhere in Geisli, suggesting its "elskig" is his attempt to make sense of a reading like B's).
Another outcome of Chase's methodology is that even where independent manuscript traditions such as SnE agree with B, their readings are rejected in favour of F (e.g. óð in 1.1 - comparison with Chase's apparatus and Skj A suggests the SnE ms. A shares this reading with B, not with the other SnE ms. W as claimed in the note on p. 123), although this only affects a few stanzas as little of Geisli is preserved elsewhere. Chase argues convincingly in favour of F at several points (e.g. 2.1, where some previous editors have emended mss heims to húms, and 11a, where all editors before him have taken B over F). But it seems wilful so insistently to prefer F when as he says, "neither of the two manuscripts of the complete text has clear advantages over the other" (4). If this were in fact the case, a "best text" edition of F might conceivably be justified. When F is less good than B, as it often appears to be from this edition, this procedure is counterproductive, as it means the text presented in the body of the book has to be withdrawn or called into question in the Commentary.
Chase's edition contains the first usable complete English translation of Geisli; the one in Corpus poeticum boreale is, as Chase observes, highly idiosyncratic. Generally it reads well, though some awkward turns of phrase ("man-ornaments" for manndýrðir, "sting" for stinga, referring to the putting out of eyes) show where the balance between literalness and idiomatic English has, as so often in translating skaldic verse, been difficult to strike. There are a few minor errors such as sg. for pl. and present tense for past, but these do not significantly affect interpretation; an exception is 66.3 where hárs must be gen. sg. of hárr "[hoary], old" not hár "high". The normalisation is quite erratic. Presumably the normalised spelling is intended to represent a reconstructed mid twelfth-century standard, although there is no discussion of the principles followed. In practice the normalised text mixes older and younger spellings, and quite a few errors have crept in in the transition from diplomatic to normalised text: o for o-caudata, i for í, é or ei for e, d for ð. In a number of cases the different elements of the edition (transcription, edited text, prose syntax, translation, and apparatus) do not tally with one another, seemingly because second thoughts or revisions appear in one place but not the others, and these problems persist into the glossary, making it a good deal less helpful than it might be. Typesetting errors are few and insignificant. Two which might cause confusion are "C" in the lexical variants to 55.6, which is not a siglum, but must be ".c.", the F scribe's abbreviation for hundrað (which appears not to be in the glossary); an "x" also seems to have dropped out of both sets of manuscript measurements on p. 4.
The preference for F even when its text is implausible and the erratic copy-editing significantly reduce the usefulness of this edition. This is regrettable considering Geisli's interest as both a well-preserved drápa and an instance of skaldic innovation in the use of Christian subject-matter and figural schemes - features Chase describes well. Chase's explication of the theological significance of Geisli and its use of Christian literary models is a great strength of his edition. His aim to "encourage... [the reader] to use the information provided in the apparatus and consider other possible solutions" (4) is laudable, but the frequent discrepancies between text and apparatus, inconsistent or absent normalisation (surely normalised texts could have been provided for the stanzas quoted on p. 26?), and sometimes minimalist notes (e.g. those to 12.5-8, 14.1-4, 40.1-4, 56.5-8) demand of the reader not just expertise but vigilance. While Chase's edition brings readers close to the late medieval text contained in F, it is debatable whether it entirely succeeds as a presentation either of Einarr Skúlason's twelfth-century poem - which must of course always be a reconstruction, though we are on firmer ground with Geisli than is often the case - or of the editor's own understanding of Geisli.