O'Donoghue starts with a nod to the notion that Old Icelandic literature belongs to an "ancient and widespread European tradition" of prosimetrum (1), but pragmatically, in accord with recent revisionist scholarship, she maintains agnosticism on whether we should posit a shared origins, even where tempting analogues, such as Irish prosimetra, lie close to hand. Her focus is upon relations between verses and prose in the sagas of Icelanders, with some limited outreach to Icelandic/Norwegian historiography. Her rationale for this focus convinces: these verses belong to a new genre, being "purposeful additions to new narratives", namely the sagas, that have "a clear terminus ante quem". Her objective is to disclose "exactly what these introduced verses contribute to the narrative--especially those which function as lyric or dialogue". (3) In a similarly synchronic gesture, what is at issue is "not the origin of the saga author's materials, but the effect that has been achieved with them..., the role of verse in the poetics of saga composition". (4)
Embedded in the discussion are, of necessity, a series of methodological and critical assumptions. It would have been good for the book if all of them could have been probed and critiqued rather more than they are. Pervasively assumed is an agent called a "saga author". To be sure, this is a prevalent (almost time-honoured) construct in scholarship and O'Donoghue is careful to qualify the meaning of the term. (13) But at the same time the assumption that an anecdote or report can only acquire polish when worked over by an "author" could with benefit have been critiqued against the recent scholarship on orality by Gisli Sigurdsson and others. O'Donoghue's methodology becomes even more markedly intuitive when to the posited author are attributed "purposeful artistry" and "conscious and self- conscious interplay of forms" (6)--these qualities being postulated side by side with "the studied artlessness of saga prose". (7) Correspondingly, the use of the term "(ideal) reader" signals an approach based in New Criticism and yet to orientate itself fully to conditions of textual production and consumption in medieval Iceland. The use of terms from current narratology is equally problematic since it carries with it the presupposition that categories that seem to us self-evident (external versus internal focalization, diegetic versus extra-diegetic, and so forth) transfer readily to medieval texts. Some twentieth-century terms appear to be misunderstood or used anachronistically. Bakhtinian "carnivalesque" seems to be equated with "wild" impulses rather than with oppositional ones. (7) "Alienation effect" is applied to describe the "reader's" reaction to finding a verse interrupting a prose narration (80), which seems a remarkable attenuation of its usual sense. Scholarly grasp of the medieval poetry likewise has its limits, for instance in the apparent unawareness that extended autobiographical poems existed (cf. p. 125) and the repeated assertion that kviduhattr is among the oldest of metres. The invocation of the genre term "lyric" hints at an imposition of foreign categories. A more welcome move is the categorization of uses of verse in prose into "documentation and dialogue" (77), which seems less tendentious than previous formulations by, e.g., Bjarni Einarsson.
In her Chapter 1, "The role of verses in Norse historical works", O'Donoghue explores the "small but significant degree of non- documentary verse function in Norse histories, which seems likely to derive not only from earlier narrative techniques in Norse story- telling, but also from the corroborative practices of learned Latin historians." (8) O'Donoghue operates here with a concept of "litteralite " (which she states she will invoke in preference to "fictionality"). She explains that "authors who present their characters as speaking their dialogue in drottkvaett are, with certain important caveats, (re)producing a textual illusion for literary effect, rather than relating a naturalistic event". (12) Hence "litteralite "--but, tellingly, it is eclipsed by more familiar but less precise expressions such as "fictionality" and even "author's creative impulse" (41) in the ensuing chapters. That is doubly unfortunate because "fictionality" is such a treacherous concept in relation to sagas, as the author herself registers. She invokes it, for instance, to dismiss saga accounts of verse speaking (137), when documentary evidence of early modern cultural practices indicates at least some factual basis. The bulk of Chapter 1 consists of O'Donoghue's minute analysis of verses and their contexts in Agrip . She contends that they are not used simply to authenticate or corroborate data given in the prose but rather to capture impressions, as with one poet's vignette of Haraldr grafeldr. The survey later broadens to Heimskringla and Legendary saga , where the development of a "prose question" + "verse answer" format is traced. Analysis of prosimetra relating to Scandinavian activities in early eleventh-century England leads O'Donoghue to the conclusion that in "coping with inconsistent source materials...Snorri is willing to elaborate, if necessary, but not to suppress". (55) Exceptions could, however, be found in his handling of Nesjavisur and verses by Torf-Einarr, to name only two examples.
In the very lengthy Chapter 2, "The community and the individual in Eyrbyggja saga ", O'Donoghue attempts to demonstrate the "mature control" of both documentary and "non-documentary" modes (NB the slippage from the term "dialogue") in this text. (9) In her view, "the complexity of the saga's structure arises from the play and tension between two directly opposed but fundamentally linked themes in the saga: the development, through time, of the community, and the place of the outstanding (or ostracized) individual within it." (80) In practice it proves difficult to sustain an argument capable (Bifrost-like) of spanning the gap between the observed facts of verse deployment and so abstract a thematization. In attempting it O'Donoghue commits herself to a teleological approach, where authorial intentions are inferred from certain features of the text. Symptomatic is the following (126): "Bjorn [Breidvikingakappi] is set apart from other members of the community, and his lofty and powerful verses mark him out as distinctively 'other' from his society. This is surely why this narrative element has been created." This statement is followed up presently by a fuller five-point analysis of such "effects" in Gisla saga (143) and an allusion to the structure of Beowulf in similar vein. (127) A practice that O'Donoghue attributes to the author of Eyrbyggja saga is the use of certain verses as "structural markers" with a "punctuating role" in the narrative (pp. 84-85). In developing this idea, in which she is anticipated by Paul Bibire among others, it is a pity that O'Donoghue lays emphasis on a tendentious value judgement that certain verses are simplistic in content. The drive to find recognizably aesthetic or "writerly" features in Eyrbyggja saga distracts from clear indications that the text is an assemblage of parts. While certainly O'Donoghue acknowledges a "patchwork" element in the saga (126), she seems not to integrate that point fully into her argument. Evidently the interpretive community held such components of the saga as the verses, the memoir of Snorri godi, and the tales of the supernatural--however empty or perfunctory they might seem nowadays-- in esteem, their reasons most probably having more to do with content than with narratology.
The next stage in O'Donoghue's exposition is a two-chapter examination of Gisla saga and Grettis saga . They are taken as contrasts to Eyrbyggja saga , since in them the verses are used in a "fully fictional, almost theatrical way, to create psychological special effects in the narrative". (9) In Chapter 3, "Speech, silence, and subjectivity in Gisla saga ", O'Donoghue argues that "the inner torment of Gisli's premonition of his own violent death is conveyed in the riddling but powerful medium of his strange, visionary poetry, and the saga author presents some of these verses as soliloquy as he explores ways of representing Gisli's subjectivity". (9; cf. pp. 160-61) The idea of subjectivity, as invoked here and elsewhere, is interesting and worthy of further investigation, perhaps in conjunction with a response to studies of mainland European developments by such a scholar as Stephanie Trigg. But in the analysis of Gisla saga it runs up against the difficulty that medieval dreams centre upon somebody or something that encounters one, or that one has encountered (hence the English word "met" in dream narratives). To that extent medieval reports of the dream "encounter" would have been perceived by contemporaries as less purely subjective, possessing more palpable reality, than is the case with dreams in modern theory. That must have applied especially when the figures within the dream had currency within contemporary systems of belief and interpretation (as is apparently the case in Gisla saga ). On the credit side, O'Donoghue's discussion of the supposed two dream- women (e.g., pp. 163 and 166) is more accurate than previous accounts and may lead to better understandings of the poetry.
O'Donoghue also seeks to demonstrate a heroic dimension for the saga and to tie it to verse-speaking. Given the well-known comparison of Gisli's sister Thordis with Gudrun Gjukadottir, such an endeavour is justified, but it can be taken too far, as for instance in the detection of "an understated echo of the actions of Gunnarr and Hogni in Atlakvida ". (151) A better opportunity for analysis of subtleties could have been found in the verse (7) where Gisli is made to compare Audr's shedding of tears with the picking of hazel-nuts. Are there covert hints of calculated display and sexual duplicity in this vignette, with its obvious mainland European analogues? Could these overtones, if present, link with other awkward moments in the characterization of Audr?
In Chapter 4, "Grettis saga and the fictionalization of biography", O'Donoghue argues that Grettir "expresses his dislocation from society through his elevated, uncompromisingly oblique skaldic discourse". (9) This case is fairly familiar from previous scholarship and does not require detailed commentary. But the assumption, quoted from Ursula Dronke, that verse is a mode of discourse "kept remote from the traffic of common speech" (183) is over-simple on several counts, as noted already, and, once again, it makes the concept of "fictionality" in relation to this aspect of saga narration suspect. In a nice piece of detailed analysis, O'Donoghue notes that the Sodulkolluvisur do not fit well with the beginning of Grettir's outlaw exploits (the context in which they are placed in the saga as we have it), and have perhaps been moved there from some other place. (200)
In a brief Epilogue, "Hrafnkels saga and the hero without verse", the author contends that "the structure of the narrative [of Hrafnkels saga ] and the characterization and dialogue of its personages are so fully developed and refined that they can carry the themes and dramatic climaxes of the narrative without the technical safety valve of verse." (9) (We might wonder at the expression "safety valve," but let that pass.) O'Donoghue's contention here, in an obverse to her case in Gisla saga , is that "it could not be part of the saga author's purposes to present Hrafnkell as a hero who speaks verses--the discourse of the heroic past". (232) This statement is made on the basis that Hrafnkell sets up his own farm rather than inheriting his father's settlement, therefore is somehow not "heroic". Quite aside from the bizarreness of such a characterization, the words "hero" and "heroic" are definitionally slippery and, as often before, the logic suffers from an excess of determinism.
The book is rounded out by a useful Bibliography and Index. It adheres to high standards of production throughout. Only a very few typographical errors appear, e.g., "predece[s]sor" (34), "Austrfara[r]visur" (67), "[the] prose" (91), "speakes" (104), "Hallger[d]r" (111), "henna[r]" (127), "br[j]ost" and "Tha", for "tha" (149), "dottir", for "doetr" (172), "corspe", for "corpse" (176), and "gengre" for "genger" (181). A puzzlingly garbled sentence is "the severed head verse speaks". (130) The translations of prose are mostly accurate but a few errors have crept in: translation of "theira", for instance, is omitted in the quotation on p. 104 and "vera vard ek nokkur" means "I had to be somewhere", not "I have". (205) The handling of the verses, although largely reliable, could have used still more care. Thus we should read "eyes...are fixed" for "standa...augu" (35), not the singular forms. "Your passionate love" as a rendition of "eitr godmunar" (165) shows vagueness about some potentially very significant vocabulary. The second helmingr of the same verse is also wrongly translated: "erlendis" is an adverb, thus "the almighty ruler of mankind has sent you alone into foreign parts from your hall to explore another world", not "from your foreign hall". (165) "Fals hallar regns Fulla" is not "goddess of the hall- spear" (what would that be?) but "goddess of battle [the rain of the hall of the top end of the spear]". (177)
The strength of this book lies in the registration of subtleties of psychology and subjectivity in the saga narratives about Thorarinn mahlidingr, Gisli Sursson, Grettir Asmundarson, and other figures. Our sense of the power of prosimetrum as a literary form will be sharpened by O'Donoghue's detailed analysis--scene by scene, dialogue by dialogue, verse by verse. In reading the verses she does not attempt a fully independent investigation of the text, but nevertheless she does refine on some previous readings, for instance in the discussion of the well-known stanzas spoken by Bjorn Breidvikingakappi (pp. 114-16). And she may well be right to see the use of verses to disclose emotions (in itself a well-known phenomenon) as marking a new cultivation of individuality and subjectivity. But that is at present merely a promising idea that awaits development through systematic theorization and presentation of evidence.