Sigvatr’s Tears: The Phenomenology of Emotion in Skaldic Verse

ORCID iD: 0000-0001-9040-5519
ISSN: Print 2754-4575
ISSN: Online 2754-4583
DOI: 10.57686/256204/22

© 2022 Gareth Evans
This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY NC)

Sigvatr, Skaldic Verse and Old Norse Emotion

In 1030, the poet Sigvatr Þórðarson hears that his beloved lord and patron, Ólafr inn digri (later inn helgi) Haraldsson, has fallen at the battle of Stiklarstaðir.¹ Like any good retainer, Sigvatr has feared his lord’s death.² His reaction to the news of Óláfr’s fall is narrated in Heimskringla and Flateyjarbók and, in both texts, we are given a series of lausavísur that express his emotional response.³ These lausavísur have occasioned a number of comments acknowledging the poignancy of their emotional expression. The latest editor of the verses, R. D. Fulk, speaks of the ‘inconsolable grief’ that Sigvatr expresses;⁴ Diana Whaley writes of the ‘intense emotional responses’ that these stanzas evidence;⁵ Jón Skaptason suggests that the poems ‘bear witness to genuine grief’;⁶ Judith Jesch refers to these verses by Sigvatr as ‘remarkably personal stanzas expressing his grief at the death of King Óláfr’ and has suggested that they ‘reveal a deeply-felt and very personal grief’;⁷ Russell Poole refers to them as ‘very eloquent and touching memorial lausavísur’;⁸ and Gabriel Turville-Petre has noted that these stanzas are ‘rich in personal feeling’.⁹ There is no shortage of scholars identifying the emotional content of these verses.

Such identifications are unsurprising given that skaldic verse has often been recognised as a vehicle for emotional expression, particularly when considered in comparison with the prose of much prosimetrical saga narrative.¹⁰ The vast majority of discussions of emotion in skaldic verse do not go beyond the mere identification of verse as a vehicle for feeling, however, and even works that explicitly take Old Norse emotion as their focus do not fully engage with skaldic verse and so do not adequately explore how its peculiar literary features might contribute to our understanding of the staging of emotion in Old Norse verse. The most substantial contribution to the study of emotion in Old Norse literature has come in the form of Sif Rikhardsdottir’s recent monograph, Emotion in Old Norse Literature.¹¹ Her work is vitally important in exploring how emotion is staged in Old Norse literature in general, but she explicitly avoids exploration of skaldic verse.¹² Likewise, Christopher Crocker’s recent chapter on ‘Emotions’ in The Routledge Research Companion to The Medieval Icelandic Sagas is indicative in that it does not even mention skaldic verse and so does not entertain how verse constructs emotional interiorities.¹³ While the study of emotion in Old Norse literature is now becoming relatively common, there is a significant lack of work on emotion in skaldic verse and, in particular, on how the complex verse form favoured by skaldic poets lends itself to emotional representation.¹⁴

Merely identifying skaldic verses as vehicles for emotion is akin to the simplifying, homogenising function of so-called ‘emotion words’ as described by much recent emotion theory (and the use of which we sometimes see in action in saga prose). Just as to give an emotion a convenient categorical label — love, hatred, grief, and so on — simplifies emotional complexity, to say that skaldic verse is ‘emotional’ without further analysis simplifies its literary complexity and function. William Reddy, for example, claims that each emotion that any one of us feels is ‘sui generis’ and suggests that ‘[s]imple emotion labels are oversimplifications.’¹⁵ Barbara H. Rosenwein and Riccardo Cristiani also point to the limitations of pigeon-holing emotions into single-word expressions.¹⁶ Jerome Kagan, too, writes of ‘the inadequacy of most languages to capture’ human emotions,¹⁷ suggesting that while ‘[f]eelings are dynamic […] semantic concepts freeze-frame these states into static categories’.¹⁸ This last quotation from Kagan also highlights the temporal, ongoing nature of emotional states, and draws on the idea of emotion as a ‘flow’, a conceptualisation which suggests that emotion is better viewed as process than isolated object or discrete event.¹⁹ Through a close analysis of a skaldic stanza, I here want to demonstrate how skaldic poets represent dynamic emotions, and also that an analysis of skaldic verse indicates — implicitly — that skaldic poets seem to have held an understanding of emotion as dynamic process. While dynamic emotions and their literary staging have been examined in Old English poetry, most notably by Antonina Harbus, Old Norse skaldic verse has not yet been explored as a form congenial to the representation of emotions in flux.²⁰ It is in this respect that this article seeks to make its contribution.

In many ways, skaldic verse goes further in exploring complex emotional interiorities than scholars who would simply identify Old Norse emotions, seemingly pre-empting Kagan’s appeal that we should ‘agree to a moratorium on the use of single words, such as fear, anger, joy, and sad, and write about emotional processes with full sentences rather than ambiguous naked concepts’.²¹ Indeed, it is my contention that skaldic verse — as a direct result of its famously complex formal features — is able to encode and convey complex, dynamic emotional interiorities in ways that move beyond the possibilities of prose.²²

Interpreting Skaldic Verse in Time

Before I continue on to primary analysis, it is necessary to consider ideas of the temporal nature of literary interpretation as this is fundamental to skaldic verse’s mode of emotional representation. Here, in focussing on the temporal experience of interpreting Old Norse literature, I take my lead from Heather O’Donoghue. In her most recent monograph, Narrative in the Icelandic Family Saga: Meanings of Time in Old Norse Literature, O’Donoghue produces a masterful analysis of time in the sagas of Icelanders, not only of how time is represented within Íslendingasaga narrative but also — and of most relevance to my analysis here — of the audience’s temporal experience of progressing through saga prose.²³ Drawing on Augustine’s discussion of the paradox of time and Paul Ricoeur’s notion of the ‘double temporality’ of narrative, O’Donoghue suggests that ‘at the same time as viewing the text from the totum simul perspective, we also live through its narrative time’.²⁴ She moreover argues that, ‘[w]e are experiencing as we read or listen a sort of recreation of Augustine’s human, impossible-to-define time — it’s rather like following the ever-moving dot on a karaoke screen.’²⁵ The karaoke dot analogy is ingenious and captures perfectly the audience’s constantly moving, dynamic focus, as they move from one narrative moment in saga prose to the next, with the current present continually receding into the now-remembered past and the audience moving forward into an as-yet-unknown-but-nevertheless-anticipated future that itself will shortly become the present moment of attention and subsequently the known past. Through this mode of analysis, O’Donoghue is able to discuss the literary effects of unfolding narrative with precision and to theorise an audience’s phenomenological interaction with saga narrative, thereby making a significant contribution to our understanding of saga poetics. But O’Donoghue’s analysis of temporal unfolding focuses on saga prose and — while she has considered the temporal effects that the inclusion of skaldic verse can have on the pacing of saga narrative — she has not explicitly considered the phenomenological and temporal experience of interpreting skaldic verse itself.²⁶ Here, I consider the phenomenological experience of reading or hearing skaldic verse, and suggest that an interpretation attuned to the temporal unfolding of skaldic poetry enables us to better understand the means by which skaldic artistry can construct and modulate emotion.

Skaldic verse is particularly appropriate to investigate in terms of how its appreciation is conditioned by the fundamentally temporal nature of the reading or listening process given that there is evidence that medieval Icelanders conceived of skaldic verse as a form intimately associated with time. Vatnsdœla saga, for example, recounts an episode that suggests that verses may have been recited according to a precise pace and that a verse could be used as a recognizable unit of time.²⁷ Moreover, although the poem under discussion in this article is not a drápa, poetic terms for the elements of the drápa, as Margaret Clunies Ross has noted, ‘lay stress on time intervals’: the word for a refrain, stef, literally means a ‘fixed period of time’, while stefjamél, means ‘an interval of time between refrains’.²⁸ Both terms reinforce the idea that skaldic verse is understood as a temporal art – or, to use Lee Hollander’s term, as one of the ‘time-arts’.²⁹ My concern in this article with the phenomenological experience of reading or listening has clear affinities with Reuben A. Brower’s method of ‘reading in slow motion’, in which he advocates ‘slowing down the process of reading to observe what is happening, in order to attend very closely to the words, their uses, and their meanings’ and also Stanley Fish’s approach in ‘Literature in the Reader’, in which his method is to ‘slow down the reading experience so that “events” one does not notice in normal time, but which do occur, are brought before our analytical attentions’.³⁰ Fish, in particular, emphasizes the importance of paying attention to the ‘temporal flow of the reading experience’.³¹ By slowing down — and critically reconstructing — the experience of interacting with a text in time we can make explicit the near-automatic interpretative moves that take place in a reader or listener’s mind and so can consider how these mental processes condition our emotional experience of, and response to, a text.³² It is this that I here undertake for a single skaldic stanza.

Sigvatr’s Lausavisa 20

The features of skaldic verse that I here want to focus upon and which I suggest are able to encode complex and dynamic emotional states are largely untranslatable and can only be appreciated through an analysis of the text in the original Old Norse. This is partly as a result of the ‘multiplicity of interpretative options’ that my chosen stanza — like many others in the same mode — presents to its readers or listeners, but also because the often convoluted syntactical relationships between words and phrases, while permissible in a skaldic stanza, cannot be represented adequately in modern English while maintaining an analogous literary effect.³³ Giving a single translation for many skaldic verses thus imposes stasis upon poetry that — in the original — is dynamic and shifting. Providing a translation, then, is inimical to the argument I seek to make. But, with this caveat in mind, I here give Fulk’s critical text of Sigvatr’s lausavísa 20,³⁴ and — for the sake of comprehensibility — his rendering of the verse into modern English.

Fúss læzk maðr, ef missir

meyjar faðms, at deyja;

-keypt es ǫ́st, ef eptir,

of-, látinn skal gráta.

En fullhugi fellir

flóttstyggr, sás varð dróttin,

várt torrek lízk verra,

vígtǫ́r, konungs ǫ́rum.³⁵

A man claims he is ready to die if he misses the embrace of a maiden; love is too dearly bought if one must weep for the departed. But the flight-shunning man full of courage who has lost his lord sheds slaying-tears; our grievous loss seems worse to the servants of the king.³⁶

The stanza — in translation, at least — appears quite straightforward. Sigvatr compares the emotion felt by a grieving man (who expresses his loss through weeping) to his own grief at the loss of his lord, which prompts the shedding of his own ‘slaying-tears’ (a compound to which I shall return shortly); Sigvatr’s weighing up of the griefs emphasises that the loss of the king is felt as greater than the loss of a woman.³⁷ What follows is a close reading of the stanza that seeks to draw out and emphasise its emotional content, largely following the order in which ideas and words are presented to us in the verse. The stanza will thus be examined in the same order that — phenomenologically speaking — a reader or listener experiences it upon their first encounter with the text: this is essential for exploring how the verse represents emotions in flux and stages the change of emotions over time.

The First Helmingr

The more dazzling emotional representation is found in the second helmingr, but the entire verse is worthy of close analysis. Beginning with the first two lines of the first helmingr, we encounter the suggestion that:

Fúss læzk maðr, ef missir

meyjar faðms, at deyja.

A man claims himself to be eager — if he misses a maiden’s embrace — to die.³⁸

The order in which we as readers or listeners receive these ideas conditions our affective response. We start with the emotional state of eagerness ascribed to an unnamed man. The state of eagerness is emphasised through the placement of the adjective fúss at the very beginning of the line; the capaciousness and indeterminacy of the eagerness is maintained when, two words later, we are then introduced to the subject modified by this adjective — maðr — translated above as ‘man’, but strictly speaking merely meaning ‘person’. We begin this stanza, then, with the reader or listener’s emotions being aroused, but without — temporarily at least — the eagerness or the desire introduced being given a target.³⁹ In this way, the stanza opens by signalling — and thus priming the reader or listener for — its emotional content but it momentarily keeps the reader or listener in a state of anticipation until the focus of this desire is revealed.

This revelation is delayed, however, by the incorporation of a sub-clause: ef missir meyjar faðms (if [he] misses the embrace of a maiden). This sub-clause introduces — with the use of missir — the idea of the loss of a person and so begins to narrow the hitherto unspecified emotional eagerness of the first clause.⁴⁰ The use of the conditional conjunction ef indicates that the emotional state being described is dependent upon this loss. The loss referenced — meyjar faðms (a maiden’s embrace), metonymically functioning for the loss of the maiden herself — not only introduces the idea of interpersonal loss that recurs throughout the stanza, but also encourages us to read an element of denigration into this reference. In much Old Norse literature, men’s sexual desire for, and activity with, women is considered a threat to masculine status (especially when contrasted to martial activity and homosocial bonds);⁴¹ this idea seems to be drawn upon here in the particular focus on the loss of the woman’s embrace, rather than the woman herself, consigning the significance of this relationship to the physical realm alone.

Following this sub-clause we learn what it is that the man is eager to do as a result of the loss of the maiden’s embrace — at deyja (to die) — and with this clarification the state of eagerness that opens the first line is given its definitive focus. We now have the completion of the stanza’s first coherent emotional states: a desire for the death of the self, prompted by the loss of a loved one (on the part of the maðr), and also — implicitly — a distaste for the man’s particular emotional response and its trigger, as suggested by the phrasing meyjar faðms.

Thus far, while the opening of the first helmingr indicates the stanza’s concern with emotional expression, and the manipulation of word and clause order to condition this emotional response, it does not yet move beyond the possibilities of emotional representation afforded by prose. The second half of this first helmingr, however, begins to demonstrate the possibilities for emotional expression peculiar to skaldic poetry. In Old Norse, these two lines run thus:

-keypt es ǫ́st, ef eptir,

of-, látinn skal gráta.

Here we seem to be told — in Fulk’s translation, at least — that ‘love is too dearly bought if one must weep for the departed’. But such a translation obscures the complexity of the syntax, which must be understood as unfolding temporally through the process of reading or listening to be appreciated fully; a single translation also precludes the possibility of multiple, shifting interpretations. The half- helmingr begins with the idea that love comes at a price: keypt es ǫ́st (bought is love). At this point, it is unclear exactly what that price might be or whether this price is acceptable (although the content of the first half- helmingr might make us inclined to assume that the price will be regarded as too great). The next clause, beginning ef eptir (if for/if afterwards), establishes a conditional relationship: that love comes at a price if a condition (to be revealed in the next line) is met. The next line begins — at least in the standard edition of the skaldic corpus — with the prefix of- (excessively/overly/too much), which is taken, by the editor of this stanza for the Skaldic Poetry Project, R. D. Fulk, to form a compound with - keypt by tmesis, leading to ofkeypt (too dearly bought).⁴² This reading of the formation of the compound by tmesis was also adopted by Finnur Jónsson and subsequently approved of by Jón Skaptason.⁴³ In this reading then, it is not until the reader or listener comes to the isolated prefix, of-, that they can clarify the emotional attitude to love here: not just that it has a price, but that that price is regarded too high.

It must be noted, however, that while Fulk’s edition neatly separates off the elements that form the compound ofkepyt through the use of hyphens, this typographical convention does not accurately reflect word division as found in the majority of the manuscript witnesses. All manuscripts containing the stanza — with the exception of Bæjarbók — have of forming a compound with the following word, látinn, giving oflátinn.⁴⁴ This does not preclude the possibility of the formation of ofkeypt by tmesis – as Anthony Faulkes has shown, transference of the first part of a compound to form a new compound with another word in spite of written word division is entirely permissible in skaldic poetry — but it might give us pause.⁴⁵ In a reading that accepts the formation of ofkeypt by tmesis, the reader or listener must mentally separate parts of a word and then rearrange them, a fact that points to the instability of any interpretation of this text. That the compound ofkeypt must be formed by tmesis through the interpretative action of the reader or listener means that two linked but nevertheless separate attitudes to love are possible here: 1) that love comes at a price, and 2) that the price of love is too great. We might also consider that the reader or listener is encouraged to move through these interpretations sequentially: first, to understand that love has a price and then, when reaching the next line, to recalibrate this attitude to one which understands love’s price as excessively high. Such ‘retrospective syntactical reanalysis’ — that depends upon an understanding of reading or listening as a process taking place in and through time — has recently been shown by Eric Weiskott to be in operation in some Old English poems;⁴⁶ approaching skaldic poetry with a similar sensitivity to the reader or listener’s phenomenological engagement with the text likewise indicates that syntactical reanalysis can reveal interpretative subtleties, and emotional ambivalences and fluctuations, in Old Norse poetry.

The opening of the last line of the first helmingr, of-látinn, can be treated in a number of ways.⁴⁷ Oflátinn, taken as a weak masculine compound noun, would mean something along the lines of ‘the ostentatious/gaudy person’⁴⁸ (in which case it refers to the lamenting man) or, alternatively, taken as a past participle, it can mean ‘the much lamented’.⁴⁹ If, however, we accept that the of of oflátinn is to form a compound with -keypt by tmesis, then we are left with látinn (meaning ‘the deceased’ or ‘the departed’).⁵⁰ The final two words of this first helmingr — skal gráta (must weep) — can be treated as unambiguous.

Bringing the various readings of the two lines together, we can see that there are a number of different possible interpretations. If we accept the tmesis of ofkeypt then the two lines would mean something like ‘love is too dearly bought if one must weep for the departed’. Here, the identity of the látinn (departed) is not entirely clear. The context and the first two lines of the stanza may well lead us to assume that it refers specifically to the maiden who has been lost. But if this were the case we would expect the feminine form (látin) rather than the masculine form (látinn), although it is true that in a minority of manuscripts (Holm. Perg. 1 fol., AM 63 fol^x, and AM 47 fol.) the feminine form is found. Given that, at the time of the poem’s composition at least, látinn and látin would not have been homophonous, and that látinn occurs in the majority of manuscript witness, it seems that this form should be preferred. If the reading of the masculine form is indeed correct then it is difficult to reconcile this with a specific reference to the lost maiden. It could instead be a reference Sigvatr’s lost lord (although its presence in the first helmingr makes this reading unlikely given the stanza’s narrative trajectory), or — more likely — a generalisation of the situation of loss, with the gendered specifics of the first two lines developing into a universalizing comment upon loss and grief in the third and fourth.

If we do not accept the tmesis, however, then — depending on how we render oflátinn — the line would mean something like ‘love has a price if the ostentatious one must weep for (his deceased wife)’ or ‘love has a price if one must weep for the much lamented’. As suggested earlier, however, this attempt to derive a ‘correct’ translation forces stasis upon a form that — arguably — is dependent for its aesthetic effects upon the ability to encourage the reader or listener to vacillate between different interpretations. Rather than viewing these interpretations as mutually exclusive, we can instead suggest that each has validity and that a reader or listener — during the process of interpretation — moves between them. The alternation between the readings ‘love has a price’ and ‘love is too dearly bought’, as suggested, stages an ambivalence over whether love is worth its cost. Similarly, the alternation between reading oflátinn as ‘the ostentatious one’ or ‘the much lamented’, enables a reader or listener’s focus to move between the grieving person (through a term which seems here to carry implicit criticism) and the object of their grief, staging the duality of grief: that its objects are both the grieving self and the person who has been lost.

In arguing for the presence of multiple, equally valid interpretations, we could alternatively suggest that the of prefix exists in an apo koinou relation with both keypt and látinn.⁵¹ Such an interpretation would be to suggest that a reader or listener could construe of with both keypt and látinn at the same time. Indeed, Roberta Frank suggests that, in such constructions, the shared word should be construed with the different clauses in which it participates ‘simultaneously’.⁵² However, following Eric Weiskott and in keeping with my stress on the phenomenological interpretation of skaldic verse in this article, I would suggest that we instead construe the different clauses sequentially and so focus on different aspects of the text in turn (thus producing different, sequential emotional responses).⁵³ In this instance, the argument is that (for the reader or listener) the of is first to be construed with keypt, and then with látinn. Such a temporally-aware reading enables and allows for the staging of ambivalence to love and the duality of grief.

The Second Helmingr

Now, to move on to the second half of the lausavísa:

En fullhugi fellir

flóttstyggr, sás varð dróttin,

várt torrek lízk verra,

vígtǫ́r, konungs ǫ́rum.

The first thing we encounter is the adversative conjunction — en — marking the movement into the second helmingr and signalling that the reader or listener is to expect some sort of contrast with the content of the first helmingr. As we will see, the opposition constructed, facilitated by the formal and typical division of the stanza into two syntactically-independent helmingar, is between an emotional other — the weeping man (who is implicitly derided) in the first helmingr and the personal emotion of the poet in the second helmingr, which, partially through this contrast, is validated.⁵⁴

Following this, we come to the noun fullhugi, translated by Fulk as ‘man full of courage’. This is a sensible rendering of the compound: Sveinbjörn Egilsson and Finnur Jónsson’s Lexicon Poeticum gives ‘uforfærdet, modig mand’ (a fearless, courageous man) and similarly the Cleasby-Vigfusson dictionary gives ‘a full gallant man, a hero without fear or blame’.⁵⁵ The compound also seems to be used in this way in a number of other skaldic stanzas.⁵⁶ But this is not the only way of construing its meaning. As Daniel Donoghue has noted, when modern day readers encounter a compound word, eye tracking studies have shown that they ‘treat the constituents of compounds as if they were separate words’ and he also notes that ‘[m]ore than a few experiments indicate that the semantic processing of a compound follows two pathways simultaneously: one that construes it as a single word […] and the other as two separate words.'⁵⁷ Which pathway is preferred dictates the dominant meaning of the compound.⁵⁸ Given that there is a space separating the compound’s constituent parts in four of the manuscript witnesses,⁵⁹ that the compound is not all that common in the skaldic corpus,⁶⁰ and that a listener (rather than a reader) will not have a visual guide as to whether the words should be understood as two simplexes or a compound, not all readers or listeners will necessarily have construed the word as meaning ‘a man full of courage’ (pathway 1) but may instead have considered the relation between the two parts of the word in more detail (pathway 2). We cannot assume that the former of Donoghue’s two pathways will be preferred consistently in construing fullhugi.

This paragraph, for sake of argument, interprets the compound preferring the latter of the two pathways. The second part of the compound, hugi, is the weak form of the strong masculine noun, hugr, with which it is used interchangeably in Old Norse sources.⁶¹ Hugr has been the focus of much discussion. As Brynja Þorgeirsdóttir has noted, hugr encompasses the range of meanings covered by the modern English words ‘“feeling” or “emotion” but exceeds those categories by including cognitive qualities, such as intelligence, will, and memory.’⁶² This capaciousness makes the term difficult to translate, but also suggests the indeterminacy of the affect being activated in the reader or listener (and, likewise, present in the man being described): just as with fúss in the first helmingr, we are primed to expect an emotion but its specificity is withheld and so the emotional arousal remains — temporarily — free-floating and undirected. This effect is heightened when we consider the first part of the compound — full — which suggests a complexity, a fullness of hugi, perhaps even an excess of a for-now-unspecified affect. The non-specificity of hugi also allows for the possibility of mixed emotions or emotional blends. Like Brynja, Judy Quinn notes that it is ‘difficult to condense the semantic range [of hugr] into anything more specific than “intense reactions”’ — which would likewise enable us to read fullhugi as priming the reader or listener for intense reaction (and to expect one from the ‘man full of hugr’) — but Quinn also further notes that ‘[t]he states of mind that constitute the full range of the word hugr do, however, seem to be those where cogitation can potentially turn into motivation, with action following’ and, similarly, that it seems to be the ‘zone where mental reaction prompts physical action’.⁶³ Thus, not only does the poet’s use of fullhugi, when understood as being construed via the second of Donoghue’s interpretive pathways, prime us to expect an affective response, but it might also encourage a reader or listener to anticipate a physical reaction from this subject.

The verb that follows — fella, here in its third-person singular present indicative form, fellir — helps us to begin to focus our interpretative efforts. Indeed, much linguistic and neurolinguistic research has indicated that when readers encounter a given verb they pre-emptively predict the verb’s object based on information they know about that specific verb.⁶⁴ The information gleaned at the verb is dependent upon a reader’s prior knowledge of how a verb can be used, in what senses and with which type and range of objects.⁶⁵ Knowing how to use a verb — and so how a contemporary reader or listener might have responded to fellir in Sigvatr’s stanza — depends upon ‘knowing its combinatory properties with other words’.⁶⁶ For verbs that have more than one sense, verbs often display a bias — known, simply enough, as ‘verb bias’ — toward one sense rather than another, and this bias can be affected by the context in which the verb occurs.⁶⁷ Such predictions have been shown to operate both in English and in case-marked Germanic languages.⁶⁸ In order to ascertain a verb’s ‘bias’, and in particular the ‘subcategorization biases for specific senses of verbs’, Mary Hare, Ken McRae, and Jeffrey L. Elman suggest the use of corpora.⁶⁹ For the analysis of the biases of a verb occurring in skaldic verse, the most appropriate corpus is the database of The Skaldic Project.⁷⁰ The new Lexicon Poeticum associated with this project lists 31 instances of fella (in its various forms) in its database.⁷¹ I list these in the appendix, and have separated the instances into different senses; the senses listed are of my devising as the entry for fella in the Lexicon has not yet been split into senses. I have augmented these with a small number of instances that are currently miscategorised in the new Lexicon Poeticum database, as detailed in the appendix. I am only here considering fella, and not the related verb falla (of which fella is the causative form).⁷² This approach results in 34 instances of fella from the skaldic corpus to be considered in establishing the bias of this verb, i.e. what a reader or listener encountering this verb is likely to infer about its object. The 34 instances occur with the following frequency for each sense:⁷³

The verbal bias is clearly towards ‘to kill’ and so, when encountering the verb fella, a reader or listener well-versed in skaldic diction is likely to predict that the verb is indicating that someone is going to be killed by the subject (this is the case in nearly two-thirds [22 out of 34] of the examples): that is, fella is likely to mean ‘to kill’ since the verb shows strong bias towards this meaning. A reader or listener may have a secondary prediction that the subject is going to ‘destroy’ or ‘damage’ something, given that this sense occurs in over a quarter of the examples listed. It is much less likely that a reader or listener will predict that the subject is merely going to ‘attack’ the as yet unknown object given the relative rarity of this sense. (It should also be noted that the first three senses listed in the above table do suggest some degree of conceptual overlap in their violent action against their grammatical object; grouping the first three senses listed in the table in this manner would likewise suggest a strong verb bias towards priming the reader or listener to expect violent action [31 out of 34 instances i.e. 91% of cases]. The senses ‘to shed (tears)’ [2 out of 34 instances] and ‘to wrap’ [1 out of 34 instances] are both rare: fella does not exhibit bias toward these senses and so it is unlikely that a reader or listener encountering fellir would predict that either of these senses are meant.⁷⁴

In the context of the stanza under discussion here it seems likely that, as a result of the verb bias exhibited by fellir, a reader or listener would assume that the fullhugi (‘man full of hugr’) is going to take violent action of some kind — in fact, that he is very likely to kill someone.⁷⁵ Such an interpretation — and indeed, such a linguistic prediction — is strengthened when we take into account the context of the utterance given that linguistic predictions are conditioned by contextual information.⁷⁶ Sigvatr, of course, has just heard that his lord, King Óláfr, has been killed in battle and so it would be a reasonable prediction to assume that he will wish to seek vengeance for his fallen patron. The context here then helps to reinforce the verbal bias toward violent action. Such a prediction is further supported by the previous two words: en sets up a contrast with the emotional outpouring of the weeping man in the first helmingr, suggesting that Sigvatr will react differently (thus also, at this point, seemingly precluding the sense of ‘to shed’), and fullhugi, as seen, either suggesting that the subject is brave (and therefore likely to take courageous action), or more broadly suggesting that ‘mental reaction’ will lead to ‘physical action’.⁷⁷

As well as constituting a linguistic prediction, we can simultaneously view this as an emotional prediction: the narrowing of our semantic predictions parallels, and is interdependent with, a narrowing of the range of our emotional expectations. Sif Rikhardsdottir has theorised that our ability to read emotion in literature is dependent upon what she terms the ‘horizon of feeling’, which ‘indicates the pre-established readerly expectations of emotional behaviour’.⁷⁸ Sif’s argument is that the ‘horizon of feeling’ sets up expectations for emotional expression based on generic commonplaces, but we might likewise see this ‘horizon of feeling’ being constructed on a more basic, verbal, level. Jerome Kagan, for example, explicitly links the experience of predicting the unknown ending of a sentence to the brain’s tendency to predict emotional states.⁷⁹ Similar to Sif’s ‘horizon of feeling’, Kagan introduces the term ‘envelope of potential feelings’: the range of potential feelings that may be evoked in a given context.⁸⁰ Here, once we have reached the end of the first line of the second helmingr, the envelope seems to narrow so that the reader or listener is led to expect a violent, retributive, emotional reaction when they encounter fellir. That fellir is in the present tense also suggests the ongoing nature of the emotional state as experienced within the lyric moment.⁸¹

The next two lines reinforce such a reading, but also serve to delay the ultimate disambiguation of fellir, which — as will be seen — only occurs in the final line of the stanza. After fellir, the reader or listener encounters flótstyggr, an adjective qualifying fullhugi, meaning ‘to be flight-shy’ i.e. ‘brave’, ‘not one to run from a fight’. This qualification further reinforces the impression gained from the previous line that we can expect violent action from the fullhugi, and so strengthens the reader or listener’s linguistic and emotional predictions.⁸²

Following this, we have two clauses that serve to further delay the progression to the object of fellir. The first of these is sás varð dróttin, meaning ‘that one who has lost his master’, further qualifying the context for, and so conditioning our response to, the man in this helmingr. Noting that he has lost his lord sets up a link (but also a significant contrast) with the man who has lost the woman’s embrace in the first helmingr. It also serves to further reinforce our prediction that the ‘man full of hugr’ is likely to take violent, vengeful action as a result of his loss. It is also worth noting that the usage of varð in this phrase to mean ‘lost’ rather ‘became’ is relatively rare;⁸³ I do not see that there is any other legitimate way to construe the phrase sás varð dróttin than that which has already been given, but the rarity of the sense might well give a reader or listener pause and so further contribute to the delay in apprehending the object of fellir.⁸⁴

Next we come to the phrase várt torrek lízk verra (our grievous loss seems worse). There are a number of elements to be noted here. The first, perhaps, is the collectivity of the emotional experience suggested: that while the man’s grief in the first helmingr is constructed as individual, here the grief over the loss of a lord is figured as shared and capacious. The word used to describe the loss itself — torrek — is also worthy of comment: usually translated as ‘grievous loss’, it is unique in the skaldic corpus (although it does occur in prose and in the title of Egill Skallagrímsson’s famed Sonatorrek). Its rarity in skaldic diction may give the reader or listener pause, causing the loss to be emphasised. We might also note that where the word torrek is used in prose, it seems to suggest not just a loss, but in particular a loss that is to be redressed, and so we may gloss torrek not simply as ‘a grievous loss’ but rather ‘a grievous cause for redress’.⁸⁵ Such a reading would again further reinforce the reader or listener’s prediction that the ‘man full of hugr’ in this helmingr is going to take violent retribution for the death of his lord. Finally, this phrase also offers the suggestion that this torrek is considered worse, presumably than that of the widower in the first helmingr, and so — through a hierarchisation of griefs — makes a claim for the primacy of the grief felt as a result of the loss of Óláfr.

It is only now, upon reaching the final line of the stanza, and having been delayed by the two clauses just discussed, that we reach the object of fellir. As I have been arguing, the stanza primes us — in multiple ways — to expect that the object will be an enemy on whom Sigvatr wreaks violent vengeance. But this is not at all what we get, and the stanza — I suggest — plays with our expectations to stage a complex emotional state in flux. The object we get is vígtǫ́r. Taken literally, the compound means ‘slaying tears’ or ‘killing tears’. Upon encountering it, the reader or listener’s first response would be to recalibrate their understanding of fellir. The predicted sense of ‘to kill’, which — as I have argued — the stanza is at pains to condition the reader or listener into expecting, is proved erroneous, and instead must be replaced with the sense ‘to shed’, as in ‘to shed tears’. In terms of O’Donoghue’s ‘karaoke dot’ analogy, the dislocated syntax upon which this effect depends disrupts, momentarily, the dot’s onward bounce. A reader or listener would at this point need to glance or think back to earlier in the stanza to confirm the presence, and reconsider the meaning, of fellir: in effect, to stretch the analogy, the karaoke dot of the moment of attention momentarily jumps back to land again on fellir before moving onwards.⁸⁶ This recalibration of our expectations stages a shift in our emotional response to the stanza and also even a momentary uncertainty over the emotion being constructed. Instead of the simulation of a vengeful wrath as we have been led to expect we instead seem to get — and progress, through uncertainty, to — a profound, and somatically-expressed, grief. The stanza, then, through the delaying of vígtǫr, stages a complex, dynamic emotional state in which anger turns into, and gives way, to grief — or perhaps, more precisely, stages a dynamic blend of grief and anger. It is because of the possibilities of a dislocated skaldic syntax — and the delaying of the verb’s object — that the stanza is able to give form to and represent the temporal experience of emotion’s dynamism. This, to my mind, is the most dazzling element of the stanza’s emotional staging.⁸⁷

The powerful effect of suspense and anticipation that can be produced by syntactical dislocation in skaldic verse has been noted by both medieval and modern commentators. Snorri, for example, discusses the form langlokur (‘late conclusions’) in Háttatal, where a sentence that is begun in the opening line of the first helmingr of a stanza is not completed until the last line of the second helmingr.⁸⁸ This is obviously more extreme than the dislocation we have here — in which a verb in the first line of a helmingr is not given its object until the last line of the same helmingr — but the term langlokur is testament to an awareness of the literary effects of syntactical dislocation. The Fourth Grammarian also recognised the rhetorical effects of delayed resolution, listing examples of syntactical dislocation under the figure of antitheton.⁸⁹ Hallvard Lie also wrote of the ‘spesielle estetiske lystfølelse’ (special aesthetic pleasure) produced by the late resolution of a clause in the last line of a stanza or helmingr.⁹⁰ Likewise, he mentions — following Konráð Gíslason — that word order can be manipulated to ‘skape sensasjon’ (shape sensation) but he does not explore precisely how this is achieved.⁹¹ With specific reference to the lausavísa under discussion in this article, O. D. Macrae-Gibson speaks of ‘the device of stress by suspension to emphasise […] ‘vígtár’’.⁹² To my knowledge, this is the only comment published that mentions specifically the delayed placement of vígtǫ́r and its effects in this stanza. What is also noticeable here, however, is that the suspense identified is not explored in detail nor is consideration given to how this delay plays with readers’ or listeners’ expectations and, as a result, their emotions. The dependence of this literary effect upon the delayed grammatical object reinforces the importance of reading or interpreting the words in a skaldic verse in their original order rather than rearranged into an artificial construction of ‘prose word order’ as is often the case in editions of skaldic verse. Anthony Faulkes, for example, suggests that he ‘prefer[s] not to re-order the words [of a skaldic stanza], since […] the word-order not only embodies the structure of the verse but also to a large extent the meaning’ and Roberta Frank writes of ‘[t]he pernicious practice of rearranging the words of a verse in prose order before attempting translation — a habit of modern scholarship that is fatal to the poetry as poetry’.⁹³ These comments can be built upon by considering the effects produced by the delaying of vígtǫ́r: not only is rearrangement into ‘prose word order’ inimical to original meaning and poetry, but also to an appreciation of a stanza’s emotionality.

The potential meanings of the compound vígt ǫ́r, and how a reader or listener might respond to this word, must also be considered more fully. The compound has been — and can be — understood in a number of ways.⁹⁴ Finnur Jónsson understands the word to mean ‘tårer over den dræbte’ (tears for the slain).⁹⁵ Lee M. Hollander likewise suggests ‘tears wept over one fallen in battle’.⁹⁶ Bjarni Aðalbjarnarson, in his edition of Heimskringla, glosses the compound as ‘tár, sprottið af vígahug’ (tears, sprung from the killing-mood), calling to mind the emotional state of víghugr associated with Víga-Glúmr of Víga-Glúms saga, in which he displays multiple extreme emotional indicia (laughter, paleness, hailstone-sized tears) when he is inclined to kill; this link suggests that Bjarni reads Sigvatr’s tears as likewise indicating his intent to kill.⁹⁷ Alison Finlay and Anthony Faulkes, in their translation of Heimskringla (which is based on Bjarni’s edition), similarly give ‘fierce tears’, with a note suggesting ‘“battle tears”, i.e. tears of rage shed by a warrior in a murderous mood’.⁹⁸ Alternatively, Ernst A. Kock gives ‘bitter tårar (blodstårar)’ (bitter tears [tears of blood]).⁹⁹ Kristen Mills offers a number of different readings and begins similarly to Kock in asserting that the compound refers to ‘blood’: in this reading, the ‘killing tears’ are drops of blood that Sigvatr will either cause his enemies to shed, presumably in seeking revenge for his dead lord, or will himself lose as he fights in memory of Óláfr.¹⁰⁰ She also suggests, however, that it is possible to read the tears as literal (salt) tears or ‘perhaps as literal tears of blood brought forth by the most intense emotional distress’, which — as she notes — is a motif present in Christian tradition.¹⁰¹ Furthermore, we could also read vígtǫ́r as ‘tears that threaten to kill or overcome the weeper’.

There does not seem to be a clear and unambiguous meaning for this word, as evidenced by the lack of critical consensus over an appropriate translation. The prose context of the stanza in Flateyjarbók is not much help in this regard in that it simply notes before the stanza that Sigvatr ‘fellde tár’ (shed tears) when he heard of the death of Óláfr (although it will clearly prime a reader of the stanza in Flateyjarbók to understand the verse’s use of fellir in the sense of ‘to shed (tears)’ and so undercut some of the emotional complexity of the stanza for those encountering it in this prosimetrical context). Following the recitation of the stanza, which Flateyjarbók claims Sigvatr composed in response to his tears being seen as unmanly, Sigvatr suggests that: ‘Uigtár kollum ver þat […] er ver fellum vid slik tidende’ (We call it slaying tears […] which we shed at such news).¹⁰² Sigvatr’s explanation, rather than disambiguating the compound, further draws attention to the question of what precisely vígtár are. The narrator’s inclusion of this attempted explanation might also suggest his own uncertainty over the compound’s denotation.¹⁰³ Sigvatr’s production of the verse in response to a suggestion that he behaves ‘okallmanliga’ (in an unmanly fashion) by crying might suggest to us that the tears are of the víghugr type, presaging violent action — or at least that that is how Sigvatr is presenting himself.¹⁰⁴ But the other men’s reading of the tears as ‘unmanly’ suggests that they read them as ‘ordinary’ tears. Different, conflicting interpretations are presented within the text, and so the tears — for the reader or listener — are inevitably an ambiguous signifier. Our inability to arrive at a single, unitary gloss is telling of the word’s inherent ambiguity and perhaps gestures towards the reason for the poet’s use of a metaphorically-loaded compound.¹⁰⁵ The word is rare. In skaldic poetry it only occurs here in this verse, and in prose it is likewise only attested in this one example from Flateyjarbók. It is very likely, therefore, that this compound, when encountered by a reader or listener, would give them pause. Upon encountering this word, given its likely unfamiliarity, a reader or listener would be forced to try to puzzle out the meaning — but such attempts, given the polysemous nature of the compound, will likely leave a reader or listener with multiple possible readings to choose from (as is evidenced by the lack of consensus noted above). Such semantic instability, with the compound encompassing meanings ranging from ‘bitter tears’ to ‘tears shed before vengeance’, neatly encapsulates a state of emotional flux or, alternatively, what is sometimes known as a ‘conceptual blend’. The word seems to suggest, at once, both intense grief and vengeful anger, or allows the movement or alternation between varied emotional states.

 The unexpected collocation of fellir with vígtǫ́r, as discussed above, at first stages a movement from vengeful anger to grief and the compound vígtǫ́r, when then considered in more detail, prompts the reader or listener to consider the blending or alternation of these states. By availing himself of the possibilities of skaldic syntax and ambiguous compounds that the skaldic form afford, Sigvatr is able to stage effectively a conflicted emotional state: both grief at the loss of a lord and patron, and fury towards his killers. By resolving fellir with the ambiguous compound vígtǫ́r, this particular skaldic verse offers grammatical resolution but denies the sense of closure, of finality — ultimately of satisfaction — that the resolution and disentangling of a skaldic stanza is sometimes thought to produce.¹⁰⁶ In this, too, it perhaps constructs and reflects a salient aspect of the emotional state staged here: that the intense and complex emotions felt at the loss of a lord do not conclude or dissipate once the stanza comes to an end. Instead, they persist beyond it.¹⁰⁷

The final two words of the stanza reinforce this reading. Here, konungs ǫ́rum, meaning ‘to the king’s servants’ or ‘to the king’s messengers’ reinforces the collectivity of the emotional response, by completing the thought begun in the previous line, and so giving: várt torrek lízk verra […] konungs ǫ́rum (our grievous loss seems worse to (or: among) the messengers of the king). The impact of the stanza’s emotional representation is thus heightened, particularly when we consider the use of the word ǫ́rum (messengers), which suggests that the emotions recounted in this stanza will continue to be recounted.

Conclusion

Having reached the end of this single stanza, it is now possible to assess the work — in retrospect — as a whole, and so precisely articulate why this verse can be seen as ‘et godt maal paa den følsomme digters sorg’ (a good measure of the sensitive poet’s grief).¹⁰⁸ Phenomenologically, it is only now that we can comprehend the stanza from the totum simul perspective,¹⁰⁹ given that — as Iser has noted — ‘it is impossible to absorb even a short text in a single moment’.¹¹⁰ From this perspective, we can appreciate the various techniques used by Sigvatr to construct emotion, particularly the dislocation of syntax, the manipulation of the audience’s expectation at the level of grammar, the use of tmesis, and the deployment of ambiguous compounds. I have demonstrated that skaldic verse is a literary art form that, through its inherent formal complexity, is able to stage and represent complex emotional interiorities. Roberta Frank has suggested that the complexity of skaldic verse ‘enables the divining of new and quite unexpected thoughts, the conjuring up of the unprecedented and the inexpressible’;¹¹¹ Frank’s appreciation of the representational possibilities of skaldic verse is perceptive and may be built upon by the argument presented in this article. Skaldic verse, as this analysis of Sigvatr’s lausavísa 20 indicates, not only enables the exploration of new and unexpected thoughts, but also facilitates the consideration of complex, fluctuating emotional states. I have also suggested the utility of a phenomenological approach, which views the skaldic text as unfolding through time (along with the reader’s or listener’s engagement) as an effective and necessary means of analysing the techniques used by skaldic poets to produce and condition emotional response. Skaldic verse, and in particular skaldic verse’s unique formal features, have hitherto not been a primary focus of studies of emotion in Old Norse literature. Although I have here had space to consider only one skaldic stanza in detail, and to explore how its poetic features enable it to stage its emotional content, this analysis demonstrates that skaldic verse should become a key focus for further work in Old Norse emotion studies.

Appendix

The following tables list instances of the verb fella in the new Lexicon Poeticum’s database, separated into senses of my own devising.

Of these, the final example listed — from Sigvatr’s Víkingarvísur, stanza 11 — although entered in the Skaldic Project’s Lexicon Poeticum under ‘ fella’ should instead clearly be listed under ‘ falla’, and so is discounted from my analysis.

There are three cases where an instance of fella has currently been miscategorised in the Skaldic Project as an instance of falla.¹¹⁶ In each of the three cases, the sense of fella is to kill:

1 I would like to thank Matthew Townend, Kate Heslop and the two anonymous reviewers for Leeds Medieval Studies for their comments on earlier drafts of this article. I have adopted many — although not all — of their suggestions. Any errors and limitations that remain are of course my own. This article is dedicated to Heather O’Donoghue. For a concise overview of Sigvatr’s life and poetry, see Russell Poole, ‘Sigvatr Þórðarson’, in Reallexikon der Germanischen Altertumskunde, ed. by Heinrich Beck, Dieter Geuenich, and Heiko Steuer (Berlin: De Gruyter, 2005), XXVIII, 382–86.

2 In his Vestrfaravísur of a few years prior to the fall of Óláfr (c. 1025–1027), Sigvatr explicitly stated that ‘konungs dauð munk kvíða’ [I will dread the death of the king]. See Sigvatr Þórðarson, ‘ Vestrfararvísur 3’, ed. by Judith Jesch, in Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), pp. 619–21 (p. 619).

3 Snorri Sturluson, Heimskringla II, ed. by Bjarni Aðalbjarnarson, Íslenzk fornrit, 27 (Reykjavik: Hið Íslenzka Fornritafélag, 1945), pp. 441–42; Snorri Sturluson, Heimskringla III, ed. by Bjarni Aðalbjarnarson, Íslenzk fornrit, 28 (Reykjavík: Hið Íslenzka Fornritafélag, 1951), pp. 14–18. Flateryjarbok: En samling af norske konge-sagaer med indskudte mindre fortællinger om begivenheder i og udenfor Norge samt annaler, ed. by C. R. Unger and Guðbrandur Vigfússon, 3 vols (Christiania: Malling, 1860–68), II, 371–72.

4 R. D. Fulk, ‘(Introduction to) Sigvatr Þórðarson, Lausavísur’, in Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), pp. 698–99 (p. 698).

5 Diana Whaley, ‘Skalds and Situational Verses in Heimskringla’, in Snorri Sturluson: Kolloquium anläßlich der 750. Wiederkehr seines Todestages, ed. by Alois Wolf (Tuebingen: Gunter Narr, 1993), pp. 245–66 (p. 250).

6 Jón Skaptason, ‘Material for an Edition and Translation of the Poems of Sigvat Þórðarson, skáld’ (unpublished doctoral dissertation, State University of New York, Stony Brook, 1983), p. 45.

7 Judith Jesch, ‘(Biography of) Sigvatr Þórðarson’, in Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), pp. 532–33 (p. 532); Judith Jesch, ‘The Once and Future King: History and Memory in Sigvatr’s Poetry on Óláfr Haraldsson’, in Along the Oral-Written Continuum: Types of Texts, Relations and their Implications, ed. by Slavica Ranković, Leidulf Melve, and Else Mundal (Turnhout: Brepols, 2010), pp. 103–17 (p. 104).

8 Russell Poole, ‘Sighvatr Þórðarson’, in Medieval Scandinavia: An Encyclopedia, ed. by Phillip Pulsiano and Kirsten Wolf (Abingdon: Routledge, 2016), pp. 580–81 (p. 580). See also: Russell Poole, ‘Sigvatr Þórðarson’, p. 384.

9 E. O. G. Turville-Petre, Scaldic Poetry (Oxford: Oxford University Press, 1976), p. 77.

10 Heather O’Donoghue, Skaldic Verse and the Poetics of Saga Narrative (Oxford: Oxford University Press, 2005); Sif Rikhardsdottir, Emotion in Old Norse Literature: Translations, Voices, Contexts (Cambridge: Brewer, 2017), esp. p. 20; Peter Foote and David M. Wilson, The Viking Achievement: The Society and Culture of Early Medieval Scandinavia (London: Sidgwick & Jackson, 1970), p. 362; Carol J. Clover, ‘Skaldic Sensibility’, Arkiv för nordisk filologi, 93 (1978), 63–81 (p. 64); Guðrún Ingólfsdóttir, ‘Um hlutverk vísna í Íslendinga sögum’, Skáldskaparmál, 1 (1990), 226–40 (pp. 227–28, 231); Margaret Clunies Ross, A History of Old Norse Poetry and Poetics (Cambridge: Brewer, 2005), pp. 66–67; Vésteinn Olason, Dialogues with the Viking Age: Narration and Representation in the Sagas of Icelanders, trans. by Andrew Wawn (Reykjavk: Heimskringla, 1998), pp. 40 and 131; Roberta Frank, Old Norse Court Poetry: The Dróttkvætt Stanza (Ithaca, NY: Cornell University Press, 1978), p. 24.

11 Sif Rikhardsdottir, Emotion in Old Norse Literature.

12 Sif Rikhardsdottir, Emotion in Old Norse Literature, pp. 4–5. For further discussion of this work’s contribution to the field of Old Norse emotion studies, see my review of the book in Speculum, 94 (2019), 1217–19.

13 Christopher Crocker, ‘Emotions’, in The Routledge Research Companion to the Medieval Icelandic Sagas, ed. by Ármann Jakobsson and Sverrir Jakobsson (London: Routledge, 2017), pp. 240–52.

14 A partial exception to this general rule may be found in the discussions of emotion in Sonatorrek. See, for example: Joseph Harris, ‘Sacrifice and Guilt in Sonatorrek’, in Studien zum Altgermanischen: Festschrift für Heinrich Bech, ed. by Heiko Uecker (Berlin: De Gruyter, 1994), pp. 173–96; Sif Rikhardsdottir, ‘Voice and Vocalisation: Sonatorrek and Eddic Poetry’, in Emotion in Old Norse Literature, pp. 79–115; Pete Sandberg, ‘ Sonatorrek: Egill Skallagrímsson’s Critique of Death’, Saga-Book of the Viking Society for Northern Research, 43 (2019), 103–24; and William Sayers, ‘Guilt, Grief, Grievance, and the Encrypted Name in Egill Skallagrímsson’s Sonatorrek’, Scandinavian Studies, 92 (2020), 229–46.

15 William Reddy, The Navigation of Feeling (Cambridge: Cambridge University Press, 2001), p. 102.

16 Barbara H. Rosenwein and Riccardo Cristiani, What is the History of Emotions? (Cambridge: Polity, 2018), p. 119.

17 Jerome Kagan, What is Emotion? History, Measures, and Meanings (New Haven, CT: Yale University Press, 2007), p. 9.

18 Kagan, What is Emotion?, p. 119. See also Jan Plamper’s discussion of Kagan’s approach to emotion in: Jan Plamper, The History of Emotions: An Introduction, trans. by Keith Tribe (Oxford: Oxford University Press, 2015), p. 277.

19 For the idea of emotion as flow, see Karl G. Heider, Landscapes of Emotion: Mapping Three Cultures of Emotion in Indonesia (Cambridge: Cambridge University Press, 1991), pp. 6–9. Heider relies perhaps too heavily on the flawed work of Paul Ekman, but his insight of emotion as ‘flow’ is an important one. Heider’s work is discussed in Jan Plamper, The History of Emotions, pp. 141–46.

20 Antonina Harbus, Cognitive Approaches to Old English Poetry (Cambridge: Brewer, 2012) — see esp. ch. 7.

21 Kagan, What is Emotion?, p. 216. This approach also has resonances with that of Stephanie Downes and Stephanie Trigg who ‘regard literary and dramatic texts as giving powerful form to emotions and feeling, as finding alternative ways of exploring affective states without merely naming them’: ‘Facing Up to the History of Emotions’, postmedieval, 8 (2017), 3–11 (p. 8).

22 For comments suggesting that skalds were fully aware of the stylistic effects that they produced through manipulation of the relatively free syntactical structures enabled by dróttkvætt, see: Hans Kuhn, ‘Die Dróttkvættstrophe als Kunstwerk’, in Festschrift für Konstantin Reichardt, ed. by Cristian Gellinek (Bern and Munich: Francke, 1969), pp. 63–72 (p. 72); Peter Hallberg, Old Icelandic Poetry: Eddic Lay and Skaldic Verse (Lincoln: University of Nebraska Press, 1975), p. 20.

23 Heather O’Donoghue, Narrative in the Icelandic Family Saga: Meanings of Time in Old Norse Literature (London: Bloomsbury, 2021).

24 O’Donoghue, Narrative in the Icelandic Family Saga, p. 11.

25 O’Donoghue, Narrative in the Icelandic Family Saga, p. 11.

26 O’Donoghue, Skaldic Verse and the Poetics of Saga Narrative, p. 6; O’Donoghue, Narrative in the Icelandic Family Saga, p. 95. For similar considerations of incorporated verses’ effect on narrative pacing, see Hallvard Magerøy, ‘Skaldestrofer som retardasjonsmiddel i islendingesogene’, in Sjötíu ritgerðir helgaðar Jakobi Benediktssyni 20 júlí 1977, ed. by Einar G. Pétursson and Jónas Kristjánsson, 2 vols (Reykjavík: Stofnun Árna Magnússonar, 1977), II, 586–99 (p. 588); Bjarne Fidjestøl, ‘Skaldenstrophen in der Sagaprosa: Bemerkungen zum Verhältnis zwischen Prosa und Poesie in der Heimskringla’, in Snorri Sturluson: Kolloquium anläßlich der 750. Wiederkehr seines Todestages, ed. by Alois Wolf (Tübingen: Gunter Narr, 1993), pp. 77–98 (p. 83); and Vésteinn Ólason, Dialogues with the Viking Age, p. 125. For more general discussion of the temporal nature of reading, and the phenomenological effects that this can produce, see: Wolfgang Iser, ‘The Reading Process: A Phenomenological Approach’, in The Implied Reader: Patterns of Communication in Prose Fiction from Bunyan to Beckett (Baltimore, MD: The Johns Hopkins University Press, 1978), pp. 274–94; Roman Ingarden, The Cognition of the Literary Work of Art, trans. by Ruth Ann Crowley and Kenneth R. Olson (Evanston, IL: Northwestern University Press, 1973), esp. ch. 2 (pp. 94–145); Paul B. Armstrong, How Literature Plays with the Brain: The Neuroscience of Reading and Art (Baltimore, MD: The Johns Hopkins University Press, 2013), esp. ch. 4 (pp. 91–130); and Stanley Fish, ‘Literature in the Reader: Affective Stylistics’, New Literary History, 2 (1970), 123–62.

27 Vatnsdœla saga, ed. by Einar Ól. Sveinsson, Íslenzk fornrit, 8 (Reykjavík: Híð Íslenzka Fornritafélag, 1939), p. 68. The relevant scene is quoted in, and discussed by, Kari Ellen Gade, The Structure of Old Norse Dróttkvætt Poetry (Ithaca, NY: Cornell University Press, 1995), pp. 26–27 and 226.

28 Clunies Ross, A History of Old Norse Poetry and Poetics, pp. 36–37.

29 Lee M. Hollander, The Skalds: A Selection of their Poems (Princeton, NJ: Princeton University Press, 1945), p. 8.

30 Reuben A. Brower, ‘Reading in Slow Motion’, in In Defense of Reading: A Reader’s Approach to Literary Criticism, ed. by Reuben A. Brower and Richard Poirier (New York, NY: Dutton, 1962), pp. 3–21 (p. 4); Fish, ‘Literature in the Reader’, p. 128.

31 Fish, ‘Literature in the Reader’, p. 128.

32 Throughout my discussion of the effects of skaldic verse on its audiences, I have used the somewhat ungainly phrase ‘reader or listener’. I have adopted this phrasing to acknowledge skaldic verse’s varied modes of reception — now, at its time of composition, and its time(s) of transmission — and also to suggest that the phenomenological arguments about the interpretation of skaldic verse that I develop in this article can be applied equally to verses that are read and those that are heard.

33 Whaley, ‘Editorial methodology’, Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), pp. xxx–xxxv (p. xxxiv).

34 The lausavísa is numbered 20 in the Skaldic Poetry of the Scandinavian Middle Ages series. It is numbered 22 in Finnur Jónsson (ed.), Den Norsk-Islandske skjaldedigtning, 4 vols (Copenhagen: Gyldendal, 1908–1915), B.I, 251. This verse is found in 18 manuscript witness (12 parchment, 6 paper) and there are thus a number a textual variants; I discuss these variants where relevant in the course of my analysis.

35 Sigvatr Þórðarson, ‘Lausavísur 20’, ed. by R. D. Fulk, in Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), pp. 725–26 (p. 725).

36 Trans. by R. D. Fulk: ‘Lausavísur 20’, in Poetry from the Kings’ Sagas 1, p. 725.

37 The prose context given in Heimskringla guides such a reading, although the basic situation is inherent in the stanza itself. Before this verse is spoken by Sigvatr, we are told that: ‘Sigvatr gekk einn dag um þorp nǫkkut ok heyrði, at einn hverr húsbóndi veinaði mjǫk, er hann hafði misst konu sinnar, barði á brjóst sér ok reif klæði af sér, grét mjǫk, segir, at hann vildi gjarna deyja’ (Heimskringla III, p. 15) [Sigvatr walked one day through a certain village and heard that some husband wailed loudly because he had lost his wife, beat his breast and tore his clothes from himself, says that he eagerly wanted to die].

38 This translation — unlike Fulk’s above — attempts to maintain the effect of the parenthetical clause.

39 Here, we might draw links to the Schachter-Singer model of emotional arousal, in which it is possible to experience a generalised emotional arousal that takes on a specific emotional meaning only when there is an external context that determines the emotional valence and tone of the arousal. See Plamper, The History of Emotions, pp. 202–03.

40 While the saga prose encourages us to read the loss of the maiden’s embrace as metonymic for her death, the stanza on its own does not explicitly suggest that the woman has died. It would also be possible, for example, to read a type of courtly lovesickness into the idea that the loss of a maiden’s embrace causes emotional pain strong enough to produce desire for one’s own death.

41 Gareth Lloyd Evans, Men and Masculinities in the Sagas of Icelanders (Oxford: Oxford University Press, 2019), pp. 98–99. See also David Ashurst, ‘Male Bedpartners and the “Intimacies of a Wife”: rekkjufélagar and vífs rúnar’, in Masculinities in Old Norse Literature, ed. by Gareth Lloyd Evans and Jessica Clare Hancock (Cambridge: Brewer, 2020), pp. 183–202.

42 Tmesis is a recognised device in skaldic poetics. Snorri possibly used the term atriðsklauf to refer to tmesis although this is not certain — see Snorri Sturluson, Edda: Háttatal, ed. by Anthony Faulkes (London: Viking Society for Northern Research, 2007), pp. 8 and 100. See also: Clunies Ross, A History of Old Norse Poetry and Poetics, pp. 110–11; and Faulkes, Poetical Inspiration in Old Norse and Old English Poetry (London: Viking Society for Northern Research, 1997), pp. 17 and 20.

43 Finnur Jónsson (ed.), Den Norsk-Islandske skjaldedigtning, B.I, 251; Jón Skaptason, ‘Material for an Edition and Translation of the Poems of Sigvat Þórðarson, skáld’, p. 325.

44 Jón Skaptason, ‘Material for an Edition and Translation of the Poems of Sigvat Þórðarson, Skáld’, p. 325; Fulk (ed.), ‘Lausavísur 20’, in Poetry from the Kings’ Sagas 1, p. 726.

45 Faulkes, Poetical Inspiration in Old Norse and Old English Poetry, p. 20.

46 Eric Weiskott, ‘Old English Poetry, Verse by Verse’, Anglo-Saxon England, 44 (2015), 95–130 (p. 99).

47 It should be noted that since ‘[i]n general, cliticisation of the definite article does not occur until the thirteenth century’ in skaldic verse (see: Kari Ellen Gade, ‘Normalisation on Metrical Grounds’, in Diana Whaley (ed.), Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), pp. xlvi–xlviii (p. xlviii)), it is very unlikely that Sigvatr would have intended or expected the construals with ofláti. However, as the manuscript evidence (discussed above) indicates, construal with ofláti is common in the manuscripts that preserve this verse. So, while for Sigvatr and his immediate audience the readings with ofláti would unlikely have occurred to them, it is worth discussing them here as readers and listeners of the verse in later centuries (from the thirteenth century onwards) may have construed the verse as involving an enclitic definite article.

48 Fulk (ed.), ‘Lausavísur 20’, p. 726; Cleasby-Vigfusson, s.v. of-láti; Lexicon Poeticum (Svenbjörn Egilsson and Finnur Jónsson), s.v. ofláti; Lexicon Poeticum (Skaldic Project), s.v. ofláti.

49 Cleasby-Vigfusson, s.v. of-látinn.

50 Cleasby-Vigfusson, s.v. látinn.

51 For discussion of apo koinou constructions in skaldic verse, see Roberta Frank, ‘Skaldic Verse’, in Old Norse-Icelandic Literature: A Critical Guide, ed. by Carol J. Clover and John Lindow (Ithaca, NY and London: Cornell Univiersity Press, 1985), pp. 157–96, (pp. 168–69).

52 Frank, ‘Skaldic Verse’, p. 169.

53 Weiskott, ‘Old English Poetry, Verse by Verse’.

54 See also: Finnbogi Guðmundsson, ‘Frá Sighvati skáldi Þórðarsyni’, Andvari, n. s. 12 (1970), 85–102 (p. 99), where he suggests that in this stanza ‘mætast tveir ólíkir heimar’ [two different worlds meet], and so draws attention to the contrast set up in this verse; Erin Goeres, The Poetics of Commemoration: Skaldic Verse and Social Memory, c. 890--1070 (Oxford: Oxford University Press, 2015), p. 132, where she also notes the importance of the connection and contrast between the helmingar; and Sven Aage Petersen, Vikinger og vikingeaand: Sighvat Thordssøn og hans skjaldskab (Copenhagen: Munksgaard, 1946), p. 122, where he notes the movement and comparison set up between griefs in this ‘overordentlig smukt opbygget Strofe’ [extraordinarily beautifully-constructed stanza].

55 Lexicon Poeticum (Sveinbjörn Egilsson and Finnur Jónsson), s.v. fullhugi; Cleasby-Vigfusson, s.v. full-hugi.

56 Lexicon Poeticum (Sveinbjörn Egilsson and Finnur Jónsson), s.v. fullhugi.

57 Daniel Donoghue, How the Anglo Saxons Read Their Poems (Philadelphia, PA: University of Pennsylvania Press, 2018), p. 135.

58 Donoghue, How the Anglo Saxons Read Their Poems, p. 135.

59 AM 61 fol. (full hugum), AM 325 VII 4^o (full huginn), GKS 1005 fol. (full huginn), and AM 45 fol. (fvll hugíN).

60 Lexicon Poeticum (Sveinbjörn Egilsson and Finnur Jónsson), s.v. fullhugi.

61 Cleasby-Vigfusson, s.v. hugi.

62 Brynja Þorgeirsdóttir, ‘The Head, the Heart, and the Breast: Bodily Conceptions of Emotion and Cognition in Old Norse Skaldic Poetry’, Viking and Medieval Scandinavia, 15 (2019), 29–64 (p. 40). The range of referents for hugr given by Snorri is testament to the complexity of this concept: see Edda: Skáldskaparmál, ed. by Anthony Faulkes (London: Viking Society for Northern Research, 1998), I, 108. This source is also cited by Brynja in her discussion of the complexity of hugr (‘The Head, the Heart, and the Breast’, p. 40).

63 Judy Quinn, ‘The “Wind of the Giantess”: Snorri Sturluson, Rudolf Meissner, and the Interpretation of Mythological Kennings along Taxonomic Lines’, Viking and Medieval Scandinavia, 8 (2012), 207–59 (pp. 230–31 and 254).

64 Gerry T. M. Altmann and Yuki Kamide, ‘Incremental Interpretation at Verbs: Restricting the Domain of Subsequent Reference’, Cognition, 73 (1999), 247–64 (p. 247). Similarly, see Todd R. Feretti, Ken MacRae, and Andrea Hatherell, ‘Integrating Verbs, Situation Schemas, and Thematic Role Concepts’, Journal of Memory and Language, 44 (2001), 516–47 (esp. p. 537), and also Michael P. Wilson and Susan M. Garnsey, ‘Making Simple Sentences Hard: Verb Bias Effects in Simple Direct Object Sentences’, Journal of Memory and Language, 60 (2009), 368–92 (p. 369).

65 Susan M. Garnsey, Neal J. Pearlmutter, Elizabeth Myers, and Melanie A. Lotocky, ‘The Contributions of Verb Bias and Plausibility to the Comprehension of Temporarily Ambiguous Sentences’, Journal of Memory and Language, 37 (1997), 58–93.

66 Mary Hare, Ken McRae, and Jeffrey Elman, ‘Sense and Structure: Meaning as a Determinant of Verb Subcategorisation Preferences’, Journal of Memory and Language, 48 (2003), 281–303 (p. 281).

67 Hare, McRae, and Elman, ‘Sense and Structure’, pp. 282 and 298; Garnsey, Pearlmutter, Myers, and Lotocky, ‘The Contributions of Verb Bias and Plausibility to the Comprehension of Temporarily Ambiguous Sentences’, p. 60.

68 Yuki Kamide, Christoph Scheepers, and Gerry T. M. Altmann, ‘Integration of Syntactic and Semantic Information in Predictive Processing: Cross-Linguistic Evidence from German and English’, Journal of Psycholinguistic Research, 32.1 (2003), 37–55. For further, and more general, discussions of the how a reader’s expectations can be set up (and subverted), see ch. 4. of Armstrong, How Literature Plays with the Brain, pp. 91–130. See also, Iser, ‘The Reading Process’, pp. 278–79; Ingarden, The Cognition of the Literary Work of Art, esp. p. 103.

69 Hare, McRae, and Elman, ‘Sense and Structure’, p. 284.

70 The Skaldic Project’s database can be accessed at <https://skaldic.abdn.ac.uk.

71 The Skaldic Project’s Lexicon Poeticum can be accessed at https://lexiconpoeticum.org/. It should be noted that the Lexicon’s data is provisional as it is a work in progress, but it is nevertheless able to provide sufficient data to elucidate verb bias.

72 These are rightly separated, although the Lexicon Poeticum, 2nd edn., seems to occasionally elide them within the entry for fella: see Lexicon Poeticum (Sveinbjörn Egilsson and Finnur Jónsson), s.v. fella. For discussion of the verb falla and its use to describe the death of warriors in battle (including some usages by Sigvatr), see Judith Jesch, Ships and Men in the Late Viking Age (Woodbridge: Boydell, 2001), p. 62.

73 In this table, the use of fella in Sigvatr’s lausavísa 20 is categorised under the fourth sense, ‘to shed (tears)’.

74 It is also possible that — upon encountering the verb fellir — a reader or listener may momentarily mistake it for the masculine noun fellir (‘slayer’). Such ambiguity would reinforce the reading of the fullhugi as likely to take violent action. See Lexicon Poeticum (Skaldic Project), s.v. fellir (noun m.).

75 The understanding of the verb’s bias toward this meaning, although it has to be critically reconstructed by a modern reader through the use of corpora (as here), would — for a native speaker of Old Norse who was well-versed in skaldic poetics — form part of what Daniel Donoghue has called a kind of ‘competence’ or ‘insider knowledge’ that contemporary readers or listeners would have automatically brought to the interpretation of poetry — see Daniel Donoghue, How the Anglo Saxons Read Their Poems, p. 8. It is worth noting that a similar analysis is not possible for fúss in line 1; the new Lexicon Poeticum suggests that speakers in skaldic verse can be fúss for a very wide range of things, and it does not appear that there is clear or coherent association of the adjective with a particular object of desire. Unlike the narrowing of meaning that fellir achieves, as argued here, fuss — until it is explicitly clarified with a subject and object – evokes a capacious and indeterminate eagerness, as argued above.

76 Hare, McRae, and Elman, ‘Sense and Structure’, pp. 282 and 298; Garnsey, Pearlmutter, Myers, and Lotocky, ‘The Contributions of Verb Bias and Plausibility to the Comprehension of Temporarily Ambiguous Sentences’, p. 60. Also see Lisa Feldman Barrett, How Emotions are Made: The Secret Life of the Brain (London: Pan Books, 2018), p. 28.

77 Quinn, ‘The “Wind of the Giantess”’, p. 254.

78 Sif Rikhardsdottir, Emotion in Old Norse Literature, pp. 18-19.

79 Kagan, What is Emotion?, p. 210.

80 Kagan, What is Emotion?, p. 5. See also Feldman Barrett, How Emotions are Made, p. 26.

81 As Ingarden has suggested, ‘The present tense is often used […] to describe, for instance, things or people. This gives them a semblance of permanence, of being beyond time, as if they were not subject to change’ (The Cognition of the Literary Work of Art, p. 125). Also see Annemari Ferreira’s discussion of the ‘presentness’ of skaldic performance in her ‘ Tíð, Tíðindi: Skaldic Verse as Performance Event’, in Performing Medieval Text, ed. by Ardis Butterfield, Henry Hope and Pauline Souleau (Cambridge: Legenda, 2017), pp. 53–69 (esp. p. 57).

82 In context, where Sigvatr was not present at the battle of Stiklarstaðir and was criticised for this absence (Heimskringla II, pp. 358, 362, 442; Heimskringa III, pp. 16–17; and Flateyjarbok, p. 372), his depiction of himself as flóttstyggr may be seen as an implicit rejection of these criticisms. His lausavísa 23 functions similarly — see Sigvatr Þórðarson, ‘Lausavísur 23’, ed. by R. D Fulk, in Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), p. 728.

83 See Lexicon Poeticum (Sveinbjörn Egilsson and Finnur Jónsson), s.v. verða.

84 After dróttin (accusative) and dróttinn (nominative) became homophonous, it would be possible for a listener (but not a reader, of course) to construe the line as sás varð dróttinn (that one who became lord), but such a reading would be mitigated against by the narrative scenario constructed by the verse. Such a misunderstanding would not have been possible at the time of the poem’s composition, at which point dróttin and dróttinn were not homophonous.

85 Of the five instances of torrek listed by the Dictionary of Old Norse Prose (s.v. torrek), four are found in law-codes. In these, torrek refers to stolen property for which redress will be sought; Sigvatr’s use of torrek may thus suggest that he has been ‘robbed’ of his lord. It may also prompt us to consider whether this verse, like Egill’s Sonatorrek, ‘kemur […] í hefndar stað’ (comes to stand in for revenge) that is desired but cannot be achieved (see Hermann Pálsson, ‘Fornfræði Egils sögu’, Skírnir, 168 (1994), 37–72 (p. 67)). The remaining instance of torrek is found in Hálfdanar saga svarta. Here, Halfdanr is at a Yule feast when all the food and ale disappears from the tables. The king tortures a Saami man, whom the king seemingly believes will be able to tell him what has happened to his feast. The king’s son, Haraldr, helps the Saami man escape and on their travels they encounter a nobleman hosting a banquet. The nobleman suggests to Haraldr that ‘Furðu mikit torrek lætr faðir þinn sér at, er ek tók vist nǫkkura frá honum í vetr, en ek mun þér þat launa með feginsǫgu’ (Your father brought about a greatly grievous loss for himself when I took some provisions from him in the winter, but I will compensate you with happy news) (Snorri Sturluson, Heimskringla I, ed. Bjarni Aðalbjarnarson, Íslenzk fornrit, 26 (Reykjavík: Hið Íslenzka Fornritafélag, 1941), p. 92). Here, the word torrek again seems to be bound up with ideas of seeking redress: the suggestion seems to be that the king treated the theft — a minor loss — as a torrek, a grievous loss to be avenged, through his actions aimed at punishing the responsible party. This reading of torrek as a loss requiring redress is reinforced by the nobleman’s suggestion that he will launa (compensate) the son for it. There is also something of a double-edged nature to the nobleman’s use of torrek here: the king inappropriately treated the minor theft as a torrek, but also — through his own death — experiences a truly grievous loss. Also see: Richard North, ‘The Pagan Inheritance of Egill’s Sonatorrek’, in Poetry in the Scandinavian Middle Ages: The Seventh International Saga Conference (Spoleto, 1988), pp. 289–300 (p. 292). As North notes, the variant form torræki also suggests ‘a loss not easily redeemed’ (p. 292).

86 We might anticipate that, if we could carry out an eye-tracking study on a medieval Icelander encountering this stanza in manuscript, we would record, when they first encounter vígtǫ́r, a saccade — specifically a regression — back to fixate on fellir before moving onwards. See Daniel Donoghue, How the Anglo-Saxons Read Their Poems, esp. ch. 4, ‘Eye Movement’ (pp. 128–54).

87 The other stanza listed by the Skaldic Project in which fella takes the sense of ‘to shed’, and collocates with tár, also evinces a similar delaying of the tears (and so perhaps a manipulation of audience response) — see Oddr kíkinaskáld, ‘ Poem about Magnús góði 2’, ed. by Kari Ellen Gade, in Poetry from the Kings’ Sagas 2: From c. 1035 to c. 1300, ed. by Kari Ellen Gade, Skaldic Poetry of the Scandinavian Middle Ages, 2 (Turnhout: Brepols, 2009), p. 33. It is also instructive to note that the Cleasby-Vigfusson dictionary (s.v. fella, B.II), indicates that fella can be used in metaphorical constructions to denote a dynamic interior state e.g. ‘falling in love’, ‘turning one’s mind to something’. The use of fella here, then, might also function as a metapoetical recognition that the stanza stages a dynamic emotion.

88 See Snorri Sturluson, Edda: Háttatal, ed. by Faulkes, p. 176.

89 The Fourth Grammatical Treatise, ed. Margaret Clunies Ross and Jonas Wellendorf (London: Viking Society for Northern Research, 2014), pp. 24–25 and 107.

90 Hallvard Lie, ‘ Natur’ og ‘Unatur’ i Skaldekunsten (Oslo, 1957), p. 33.

91 Lie, ‘Natur’ og ‘Unatur’ i Skaldekunsten, p. 3. See also Turville-Petre, Skaldic Poetry, pp. lx-lxi, where he notes the intentionality behind the effects produced by syntactical dislocation.

92 O. D. Macrae-Gibson, ‘Sagas, Snorri, and the Literary Criticism of Scaldic Verse’, in Úr Dölum til Dala: Guðbrandur Vigfússon Centenary Essays, ed. by Rory McTurk and Andrew Wawn (Leeds, 1989), pp. 165–86 (p. 167).

93 Faulkes, Poetical Inspiration in Old Norse and Old English Poetry, p. 11; Frank, Old Norse Court Poetry, p. 11.

94 For examples and discussion see R. D. Fulk’s notes to his edition of the stanza: Sigvatr Þórðarson, ‘Lausavísur 20’, pp. 725–26. Also see the discussion of different possible readings for this word in Árni Davíð Magnússon’s MA thesis ‘Sem hagl er höfðingjans tár: Blóðrauð hagltár íslenskra miðaldabókmennta’ (Háskóli Íslands, 2020), p. 21.

95 Finnur Jónsson (ed.), Den Norsk-Islandske skjaldedigtning, B.I, 251.

96 Lee M. Hollander, ‘Sigvatr Thordson and his Poetry’, Scandinavian Studies and Notes, 16.2 (1940), 43–67 (p. 57).

97 Bjarni Aðalbjarnarson (ed.), Heimskringla III, p. 15; Víga-Glúms saga, in Eyfrirðinga sǫgur, ed. by Jónas Kristjánsson, Íslenzk fornrit, 9 (Reykjavík: Hið Íslenzka Fornritafélag, 1956), pp. 1–98 (p. 26).

98 Snorri Sturluson, Heimskringla: Volume II, trans. by Alison Finlay and Anthony Faulkes (London: Viking Society for Northern Research, 2014), p. 300; Snorri Sturluson, Heimskringla: Volume III, trans. by Alison Finlay and Anthony Faulkes (London: Viking Society for Nothern Research, 2015), p. 10.

99 Ernst A. Kock, Notationes Norroenae, §1120. Also see: Ernst A. Kock, ‘Old West Germanic and Old Norse’, in Studies in Philology: A Miscellany in Honor of Frederick Klaeber, ed. by Kemp Malone and Martin B. Ruud (Minneapolis, MN: The University of Minnesota Press, 1929), pp. 14–20 (p. 17). We might also be tempted to associate such a reading with ‘blood rain’, ‘a widespread classical and medieval topos’ that is ‘portentous in family saga narrative’ (O’Donoghue, Narrative in the Icelandic Family Saga, p. 123). Such a resonance would likely encourage a reader or listener to interpret the vígtár as an ill omen. We could further read the blood tears as a symptom of humoral excess, with the tears as a means of purging the body of this excess, following the ‘hydraulic model’ of emotion. For evidence that humoral thinking circulated in medieval Iceland, see Brynja Þorgeirsdóttir, ‘Humoral Theory in the Medieval North: An Old Norse Translation of Epistula Vindiciani in Hauksbók’, Gripla, 29 (2018), 35–66. For the idea of tears as purgative in some medical theories, see Rosenwein and Cristiani, What is the History of Emotions?, p. 73.

100 Kristen Mills, ‘Grief, Gender, and Genre: Male Weeping in Snorri’s Account of Baldr’s Death, Kings’ Sagas, and Gesta Danorum’, The Journal of English and Germanic Philology, 113 (2014), 472–96 (p. 484). Also see Goeres, The Poetics of Commemoration, p. 133.

101 Mills, ‘Grief, Gender, and Genre’, p. 484.

102 Flateyjarbok, II, 371 and 372.

103 Goeres, The Poetics of Commemoration, pp. 133–34.

104 Flateyjarbok, II, 371.

105 Indeed, as Clover has suggested, ‘[t]hat skalds, with their well-known fondness for word-play, should also pursue deliberate ambiguities on the metaphoric level comes as no surprise’ (Clover, ‘Skaldic Sensibility’, p. 73). Also see Armstrong, How Literature Plays with the Brain, p. 87, for discussion of the effects of metaphor on the reader.

106 Lie, ‘ Natur’ og ‘Unatur’, p. 33; Frank, ‘Skaldic Verse’, p. 183; Macrae-Gibson, ‘Sagas, Snorri, and the Literary Criticism of Scaldic Verse’, p. 178.

107 There is not space here to consider fully the complex temporal effects produced by the integration of this stanza into its prose contexts, but in this footnote I offer some preliminary thoughts. Heather O’Donoghue and others have considered the ability of skaldic verse within prosimetrical saga narrative to slow the pace of narrative progression (O’Donoghue, Skaldic Verse and the Poetics of Saga Narrative, p. 6; Magerøy, ‘Skaldestrofer som retardasjonsmiddel i islendingesogene’, p. 588; Fidjestøl, ‘Skaldenstrophen in der Sagaprosa’, p. 83; Vésteinn Ólason, Dialogues with the Viking Age, p. 125). We might also consider skaldic verse as an example of isochrony within saga narrative (a form in which there is normally ‘disproportion […] between the time events would take in a real world, and the time the narrator takes to narrate them’ (O’Donoghue, Narrative in the Icelandic Family Saga, p. 13); in isochrony, a narrative ‘comes close to taking the same amount of time to relate — that is, to actually or mentally enunciate — as the events themselves might have taken to unfold’ (O’Donoghue, Narrative in the Icelandic Family Saga, p. 64). As O’Donoghue notes, isochrony is a distinguishing feature of what Gérard Genette has termed a ‘scene’ (O’Donoghue, Narrative in the Icelandic Family Saga, p. 64). If we consider Sigvatr’s lausavísa 20 in this way — as slowing the narrative progression and creating a scene, in which there is a noticeable temporal shift from the rest of the narrative — then we may consider that this stanza constructs and stages a relatively-prolonged moment of grief (or, more accurately, of the complex emotional blend produced by this verse), in which the emotion becomes more noticeable, more poignant, and more extended. In effect, as a result of the temporal shift from prose to verse, we are forced to dwell and linger within the emotion produced by the skaldic stanza. We might also consider the temporal complexity attendant upon the interpretation of skaldic verse. The inherent complexity of skaldic verse means that it is not always possible for comprehension of a verse to be immediate (or, indeed, ultimately conclusive) — for examples and discussion of delayed comprehension of skaldic verse, see Klaus von See, Skaldendichtung, p. 97; Gade, The Structure of Old Norse Dróttkvætt Poetry, pp. 23–24; and Clunies Ross, A History of Old Norse Poetry and Poetics, pp. 65–66. That interpretation and consideration of a verse can extend beyond the moment of first reading or hearing means that the emotion encoded in a given verse can be endowed with temporal extension beyond the isochronic moment of its telling, and can allow the emotion to resonate beyond the end of the stanza and into the surrounding narrative. In this regard, it is also worth noting that the verse appears twice in Heimskringla (Heimskringla II, pp. 441–42; Snorri Sturluson, Heimskringla III, pp. 14–18): encountering the verse upon its second occurrence means that the reader revisits the emotions of loss produced by the stanza, the repetition of which stages the lingering, reoccurring nature of grief and related emotions. It should also be noted that although the stanza stages a complex interior state encompassing, among other emotions, grief, the stanza is able produce in the reader or listener a concurrent emotion of aesthetic pleasure derived from the stanza’s fine literary artistry. The former type of emotion is a ‘Fiction-based emotion’ — an emotion that represents the emotion of characters within the narrative world or a response to them; the latter is an ‘Artefact-based emotion’ — an emotional response to the textual construction of the scene and the reader’s awareness of this construction. For this distinction, see Ed S.-H. Tan, ‘Film-Induced Affect as a Witness Emotion’, Poetics, 23 (1994), 7–32 (esp. p. 13).

108 Fredrik Paasche, ‘Sigvat Tordssøn: Et Skaldeportræt’, Edda, 8 (1917), 57–86 (p. 77).

109 For discussion of the Boethian totum simul perspective, see: O’Donoghue, Narrative in the Icelandic Family Saga, pp. 10–12.

110 Iser, ‘The Reading Process’, p. 280. For a similar formulation to Iser’s, that is specifically written in relation to saga prose (although not verse), see Carl Phelpstead, ‘Time’, in The Routledge Research Companion to the Medieval Icelandic Sagas, ed. by Ármann Jakobsson and Sverrir Jakobsson (London, 2017), pp. 187–97 (p. 190): ‘There can be no narrative without time: no sequence of events to be told and no sequential telling of them.’

111 Frank, Old Norse Court Poetry, p. 29.

112 Quotations and translations in these tables are taken (and, in the case of the translations, occasionally slightly modified) from those published in the Skaldic Poetry of the Scandinavian Middle Ages series.

113 As Kari Ellen Gade notes, ‘the object of this verb is likely to have been some type of wooden structure that was destroyed by the fire’ — see Grani skáld, ‘ Fragment 1’, ed. by Kari Ellen Gade, in Poetry from Treatises on Poetics, ed. by Kari Ellen Gade and Edith Marold, Skaldic Poetry of the Scandinavian Middle Ages, 3 (Turnhout: Brepols, 2017), p. 196.

114 The precise sense of this stanza is somewhat elusive. See Rolf Stavnem’s discussion in the notes to his edition of the text: Hallar-Steinn, ‘ Rekstefja 22’, ed. by Rolf Stavnem, in Poetry from the Kings’ Sagas 1: From Mythical Times to c. 1035, ed. by Diana Whaley, Skaldic Poetry of the Scandinavian Middle Ages, 1 (Turnhout: Brepols, 2012), p. 922. It seems from the prose context that the saga author understood ‘fella […] aska’ to mean that Eiríkr was ordering timbers to be thrown at Ormr. The stanza’s wording could also suggest attacking with spears or ships.

115 The context here is that Án and his adversary, Bjǫrn inn sterki, wrestle and throw each other into a fire. See Beatrice La Farge’s notes to her edition for further discussion: ‘ Áns saga bogsveigis 1 (Án bogsveigir, Lausavísur 1)’, ed. by Beatrice La Farge, in Poetry in fornaldarsögur, ed. by Margaret Clunies Ross, Skaldic Poetry of the Scandinavian Middle Ages, 8 (Turnhout: Brepols, 2017), p. 4.

116 I want to stress here again that the Skaldic Project’s website is a work in progress, so such errors are inevitable. Even with the occasional error it is nevertheless a vital resource for the analysis of skaldic verse.