Medieval multilingualism and the expression of emotion: fear in the Gawain-poet's texts

Abstract

The Gersum Project has significantly developed our understanding of Norse-derived terms in English by providing a highly systematic typology for their identification. However, this article shows that, in order to fully comprehend the lexical impact that Anglo-Scandinavian contact had on medieval English, we need to go beyond the identification of the Norse-derived terms and explore their process of integration into English. To exemplify the benefits of this approach, the article analyses the make-up of the lexico-semantic field of emotion, particularly fear, in the texts attributed to the Gawain-poet, and examines the interaction between native, Norse- and French-derived terms. This analysis moves away from the traditional study of the texts’ vocabulary in relation to their sociohistorical context, considering the terms instead from semasiological, onomasiological and stylistic perspectives. By taking this novel approach, this article addresses key linguistic and literary topics: the formal and semantic factors that facilitated the integration of Norse-derived terms into this lexico-semantic field and, more broadly, the impact that multilingualism had on the expression of emotions in medieval England; diachronic and diatopic variation in the field; and the Gawain-poet's artistry and interest in fear as a key emotion closely linked to other affective and cognitive processes.

1 Introduction

Old Norse was an important source of loans for the vocabulary of medieval English, particularly as a result of Anglo-Scandinavian contacts before the Norman Conquest. The introduction of Norse-derived terms was most probably facilitated by the similarity between Old Norse and Old English, but the close connections between the languages also significantly complicate the process of identification of Norse-derived terms in English.

The Gersum Project: The Scandinavian Influence on English Vocabulary, which I co-lead with Richard Dance and Brittany Schorn, has aimed to revolutionize our understanding of the Norse-derived terms attested in English. Its main output is a freely available database with etymological, dialectal, textual and semantic information for the terms that have been attributed a Norse origin by previous scholars and are recorded in a group of late Middle English texts from the Scandinavianized North and North-West Midlands associated with the so-called ‘Alliterative Revival’ (over 900 terms in total). At the core of the database lies a typology, piloted by Dance (2019), where the terms are divided into four main categories (A–D) and various subcategories according to the different types and levels of reliability of the evidence that we have for their identification as Norse-derived. This highly systematic typology makes a very important contribution to the field of English lexicology because it substantially increases consistency in etymological decisions.²

Nonetheless, identifying the Norse-derived terms is just the first step. In order to establish their impact on the vocabulary of medieval English, the discussion necessarily has to move next to their process of integration into and accommodation to English, in terms of their sociolinguistic (dialectal and sociolectal) distribution and, closely linked with this, the semantic and stylistic relationships that the terms established with their near-synonyms. This step is equally difficult because our research is necessarily restricted by our limited data and lack of direct access to contemporary informants. However, in spite of the complications, this is a key part in our understanding of how and why these terms were used, and there is much that can be learned from the careful analysis of the make-up of particular areas of vocabulary and intertextual comparisons, as this article hopes to demonstrate.

This article examines the integration of Norse-derived terms into the lexico-semantic field of emotion (particularly the subfield of fear) in the idiolect of the fourteenth-century Gawain-poet. His texts (Sir Gawain and the Green Knight, hereafter SGGK, Pearl, Cleanness and Patience), from Cheshire or nearby, are key members of the Gersum corpus. By focusing on a (sub)field which also attracted a high number of French loans, the work carried out here is significant beyond the article's immediate focus as this enquiry throws further light on the need to consider the complex multilinguistic situation in medieval England. Thus, the argument moves beyond artificial boundaries established in some scholarly research around the interaction between English and a single other language in order to explore the functioning of a lexico-semantic (sub)field as a whole, and provides further evidence for the need to take a nuanced approach to the impact of multilingualism on the lexis of medieval English. After an explanation of various methodological decisions (section 2), the discussion focuses on the formal and semantic factors that facilitated the poet's lexical choices (sections 3.1 and 3.2, respectively), as an indication of some of the issues at play in the integration of the terms into their lexico-semantic subfield.

2 Focus and methodological decisions

There are various factors that make the study of the vocabulary of emotion significant for our understanding of the impact that language contact had on the development of medieval English lexis in general and the poet's idiolect in particular:

(i) From a linguistic perspective, this lexico-semantic field is a central component of a language's expressive vocabulary, which is in constant need of new members to transmit a wide range of nuances, i.e. it is one of Weinreich's (1968: 58) ‘onomastic low-pressure areas’. Indeed, there was much change amongst the terms for emotion from the Old to the Middle English period: over 2,000 of the Old English terms in this field became obsolete early in the Middle English period and an almost equal number were first recorded during the fourteenth and fifteenth centuries (Diller 2014: 114).³ Given the impact of language contact on the development of this lexico-semantic field, there have been recent calls to bring together the findings in the study of historical emotions and medieval multilingualism in order to gain a better understanding of ‘the range of emotional utterance in Middle English texts’ (Downes & McNamara 2016: 451).

(ii) From formal and stylistic perspectives, the general need for varied expressive vocabulary was heightened in these four texts and, in general, all the texts associated with the ‘Alliterative Revival’ because their metrical structure required the use of near-synonyms starting with a different sound (the poet also had to contend with rhyme in SGGK and Pearl). This, as well as the poets’ interest in technical detail, has made these poems (particularly those from the North Midlands and the North) well known for their extremely rich vocabulary, with words of different etymologies, besides varied dialectal and sociolectal associations, coming together in a single text (see Turville-Petre 2018: ch. 2). Thus, these texts are extremely helpful for exploring ‘the range of emotional utterance in Middle English texts’, in Downes & McNamara's words.

(iii) From a thematic perspective, the Gawain-poet has been identified as someone highly interested in ‘psychology, and in affective and cognitive processes’ (Saunders 2015: 40), with this concern regarding the interplay between feelings and thoughts being reflected in all four texts, despite the notable difference between the chivalric topic of SGGK and the homiletic focus of the other three texts.

There is, then, much that commends the study of the lexico-semantic field of emotion in the Gawain-poet's texts but this cannot be undertaken without giving due consideration to a number of factors. We first need to delimit the field, which entails exploring what medieval speakers would have understood as an emotion. After all, the term emotion itself was first recorded during the Early Modern English period and it initially referred to political and civil unrest; it did not acquire its current meaning until the seventeenth century (OED, s.v. emotion, n.). Middle English speakers used other terms (e.g. ME mōd, affect and passioun) but they are not full synonyms of emotion (cf. Kiricsi 2005). The medieval terms suggest the presence of external forces that affect the individual. Diller (2007) explains that the transition from passioun to emotion reflected a change from a medieval God-centred worldview to a modern anthropocentric worldview, where the focus lies on the person experiencing the emotion, not on the external forces affecting him/her. Nowadays, a comprehensive definition of emotion is still disputed and this difficulty is exacerbated when it comes to historical works (see Diller 2014: ch. 1). One solution would be to look at contemporary works focusing on emotions and their (near-)equivalents to understand, generally speaking, what speakers conceived of as an emotion. This is what Rosenwein (2016) does in her study of emotions and their expression by various emotional communities (understood as groups of people who share similar emotional values and express them according to similar social norms) from 600 to 1700. She explores the feelings that contemporary writers associated with the perceived seat of emotions (e.g. Latin animus/anima and Middle English [ME] herte), the emphasis being on the feelings themselves rather than their lexical expression. While this approach works very well in a sociohistorical study like Rosenwein's, it is not equally helpful when the emphasis lies on the identification of the terms that form a lexico-semantic field, which is the focus of this article. Accordingly, I follow instead the lexico-semantic classification in The historical thesaurus of English (hereafter HTE), the best available tool for the study of English historical semantics. Admittedly, this decision does not come without problems either, for we need to be clear that, when using the HTE, we are adopting a modern typology which might not fully coincide with medieval categorizations.

Closely linked with this, it is important to consider that ‘emotions are an area of medieval life and thought where many discourses overlap, including the traditions of virtue and vice, of sin, of physiological, psychological and medical studies, and class- and gender-based ideologies’ (Lynch 2015: 49). Indeed, when exploring the expression and conception of an emotion in medieval times, it is necessary to understand a wide range of issues, such as the role of the Church (for instance, the impact of St Thomas Aquinas’ views on emotions in his Summa theologiae, 1265–74); the influence of Galen's (129 – c.200) psychosomatic humoral theory; or the different responses to emotions in the various emotional communities.⁴ It is therefore very important not to assume that modern and medieval conceptualizations of an emotion can be unproblematically equated. In fact, awareness of the differences between past and present understandings of emotions has led to much recent interest from cognitive linguists, precisely because of their belief in the inseparability between cognition and language (see Tissari 2017). The approach in this article is based mainly on structuralist semantics instead (bringing together semasiology and onomasiology to investigate the poet's lexical choices), but, where relevant, references to salient cognitive metaphors are also made.

The HTE, the Middle English Dictionary (hereafter MED), and the editions of the four texts attributed to the Gawain-poet by Andrew & Waldron (2007), and Putter & Stokes (2014) have been the guiding tools for the identification and classification of the terms belonging to the lexico-semantic field of emotion (HTE, 2.4) on the basis of the careful reading of the texts. The notes and glossaries in the editions have been extremely helpful for establishing the meaning of a word in a particular context, given that, not only is emotion as a category fuzzy but also the various emotions that we currently identify can themselves be fuzzy, with terms linking them through interesting cases of polysemy (see below section 3.2). While it is important to establish the main meaning of a word in a particular context, this article also demonstrates that we have to bear in mind the semantic range of a term when analysing the reasons behind its use and what they can tell us about its place in its lexico-semantic (sub)field.

The appendix at the end of the article presents the list and classification of terms referring to emotion in the four texts on the basis of the aforementioned resources. The etymological information provided there and in the table below, which gives a summary of the distribution of the terms in relation to their origin, relies on the information in the OED (for the native and French-derived terms) and the Gersum database (for the Norse-derived terms).⁵ As far as the latter are concerned, terms classified as A–C (without doubling or tripling of the consonant) are given as certainly or very likely to be Norse-derived, and the remainder as simply possibly or even unlikely to be Norse-derived. The phrases Norse- and French-derived terms are preferred to Norse and French loans because the former comprise not only loans but also new-formations created on the basis of loans. Morphological productivity is a sign of the integration of a loan into its recipient language, and bringing loans and new-formations together gives a better idea about the ultimate impact of multilingualism on the field.

Table 1 shows that, together with native words, French-derived terms make up a very substantial part of the field. This might seem somewhat surprising given that such terms have traditionally been associated with the vocabulary referring to the social activities where (Anglo-)French had particular importance in sociolinguistic terms: e.g. administration, government, law, religion, and the courtly and leisurely pursuits of the higher social classes. Indeed, this has been the narrow focus of studies that have looked specifically at the use of French-derived terms in the Gawain-poet's texts (e.g. Clough 1985; Volkonsaya 2013). These works highlight the overt sociolinguistic prestige of the terms and the connotations of refinement that they bring with them. However, recent studies have contributed to change such perceptions: e.g. The Bilingual Thesaurus of Everyday Life in Medieval England compiled by Richard Ingham and Louise Sylvester, where one can see the significant impact that French-derived terms had on occupational domains such as shipping and metalworking; and, more importantly for our purposes, Ingham's (2018) work in relation to the role of the clergy in the transmission of vocabulary beyond the lexico-semantic fields that have traditionally been associated with the higher, French-speaking classes. In this respect, the proportion of French-derived terms in the subfields of compassion and pride, which would have often been discussed in religious settings, is particularly notable.

Table 1. Etymological distribution of the terms for emotion in the Gawain-poet's texts

The important presence of Norse-derived terms in the field of emotion is in keeping with the fact that most Norse-derived terms are associated with non-technical fields, an indication of the close social interaction between speakers of English and Norse. Regarding their distribution, their prominence in the subfield of humility stands out, with the ME mēk and loue word-fields being core members. This is not surprising, given that the dominance of the former is already visible in the twelfth-century text Ormulum (most probably from south Lincolnshire; see Pons-Sanz 2015). The subfields of anger and fear are also worth exploring further because of the interesting amalgamation of native, French- and Norse-derived words. The expression of anger during the Middle English period has already attracted much attention from historical linguists in recent years (e.g. Gevaert 2007; Diller: 2014: ch. 14). That is not the case for fear,⁶ even though its centrality in the emotional life of medieval England and its thematic representation in literary texts have been studied in some detail (e.g. McCann & McKechnie-Mason 2018). Notably, Johnson (2000) demonstrates that the Gawain-poet had a particular interest in this emotion, and exploited in his texts the potential of rhetorical and didactic discourses of the modus timendi in order to ‘inspire in his audience the desire to flee from sin and damnation and approach fear-inspired, reverent perfection’ (2000: v; see also Wallace 1991). Johnson, who does not pay any attention to the actual terms used for the expression of this emotion, focuses only on Patience and Cleanness, as the analysis of these two poems can take advantage of the fact that they are based on biblical texts, and a comparison between the poems and their sources can elucidate how the poet wove into them his discourse of fear (cf. Putter 1996: chs. 3 and 5). Nonetheless, the other two poems attributed to him (particularly SGGK) also reflect his deep interest in this emotion. Thus, given the significance of fear from a lexical and a thematic perspective, the remainder of this article explores the interaction between native and non-native terms belonging to this lexico-semantic subfield in the Gawain-poet's texts. While the words whose Norse origin is deemed to be certain or very likely (viz. ME auelī, rade, skerren and uglī) receive particular attention, the article presents an overview of the subfield and touches on most of the terms recorded in the four poems so as to explore more widely the impact of multilingualism in the expression of fear.

3 The lexico-semantic subfield of fear in the Gawain-poet's works

Before the discussion focuses on the way in which the terms referring to fear function in the texts from a semantic and stylistic perspective, it is important to address the two main caveats identified in section 2 in relation to the study of historical emotions. The focus on fear as an important affective factor in this article is justified not only by the HTE's (2.4.21) inclusion of this lexico-semantic subfield in the more general field of emotion but also (and in terms of Rosenwein's 2016 approach) by the poet's own association of this feeling with the heart (e.g. Cleanness, l. 1723; Patience, ll. 367–8).⁷

More importantly, it is necessary to understand how this particular emotion was conceptualized by the poet and his contemporaries. Johnson (2000: chs. 1–2) and Loughlin (2002) explain that the late medieval understanding of fear was greatly influenced by the discussions in well-known philosophical and theological works, such as Peter Lombard's Sententiae (c.1159) and Aquinas’ Summa theologiae.⁸ These works identified different types of fear, which were broadly classified as follows:

1. (i) Morally neutral fear (timor naturalis): this is the inherent fear of harm or death that both human beings and animals experience (e.g. ‘Der drof in þe dale, doted for drede’, SGGK, l. 1151; ‘the deer hurtled into the valley, frenzied with fear’).

2. (ii) Sinful or culpable fear (timor libidinosus): as it involves human cognition and will, it is a perversion of timor naturalis. Its two main subtypes are timor humanus (inordinate fear of death and bodily harm, to the extent that it leads one to make morally unacceptable decisions) and timor mundanus (inordinate fear of losing material possessions). While the Green Knight's reproach that ‘al dares for drede withoute dynt schewed’ (SGGK, l. 315; ‘everyone is cowering in fear without a blow being offered’) represents an example of the former, the jeweller's fear of losing his child, his pearl, forever (e.g. ‘I dred onende quat schulde byfalle / Lest ho me eschaped þat I þer chos’, Pearl, ll. 186–7; ‘I was afraid about what might happen, in case she whom I beheld there eluded me’) could be understood as an example of the latter because it prevents him for quite a long time from understanding what she is trying to teach him. In SGGK the poet is interested in the close connection between timor naturalis, humanus and mundanus: even though the Green Knight takes Gawain's decision not to hand in the girdle as a direct consequence of the fact that the latter ‘lufed [his] lyf’ (SGGK, l. 2368; ‘loved [his] life’), Gawain interprets it in moral terms, linking various moral failings (including ME couardīse ‘cowardice’ and coveitīse ‘greed’) together (see SGGK, ll. 2374, 2379–80 and 2508; see further below, section 3.2).

3. (iii) Laudable fear (timor gratuitus): here timor naturalis is elevated for spiritual purposes. Particularly important are timor servilis (it replaces fear of physical death with fear of spiritual and eternal death, following God's judgement and punishment) and timor filialis (this is fear in its most perfect sense, as it refers to one's fear of God the Father because of one's pure love for him and one's recognition of his superiority). The texts also record examples of these subtypes: on the one hand, the inhabitants of Nineveh change their ways because of timor servilis after Jonah has finally delivered God's message (‘Such a hidor hem hent and a hatel drede, / Þat al chaunged her chere and chylled at þe hert’, Patience, ll. 367–8; ‘such a [hidor] seized them and a cruel dread, that their demeanour all changed and they grew cold at the heart’).⁹ On the other, Noah is presented as the perfect example of someone whose life is dictated by timor filialis (‘In þe drede of Dryȝtyn his dayez he vsez’, Cleanness, l. 295; ‘in the fear of God he spends his days’; cf. Cleanness, l. 342).

The quotations above exemplify that the Gawain-poet was fully aware of the different types of fear discussed in near-contemporary works and surveyed their different effects on one's behaviour, and their moral and theological implications. More importantly for our purposes, they also show that he did not attempt to distinguish between them lexically, for the native ME drēde ‘dread’ word-field is used in connection with all of them. Indeed, the members of this word-field, together with those of the ME fēre ‘fear’ word-field, can be identified as the core terms of the lexico-semantic subfield, as suggested by the number of their attestations and, particularly in the case of ME drēde, its low alliterative rank (cf. St Erkenwald, l. 233).¹⁰ In this respect, we see here (as in other Middle English texts; cf. Ogura 2013: 68–74) a clear difference from the expression of fear in Old English, when the members of OE ēge ‘fear’ word-field were the core terms (Díaz Vera 2011: 88–90; see further below, section 3.2.1).

Besides the different types of fear, the Gawain-poet was keen to express particular nuances in relation to the conceptualization of this emotion and its links to other affective and cognitive processes. Other members of the lexico-semantic subfield, most of them Norse- or French-derived terms, were especially helpful in this respect and it is to these lexical choices that we now turn our attention. The remarks below first assess the impact of the poem's metrical structure on the use of some Norse-derived terms before the discussion focuses on semantic matters.

3.1 Alliteration and the choice of terms for fear

It is widely acknowledged that sometimes formal rather than semantic factors (particularly alliterative needs) are likely to have been the main driving forces behind the selection of a term instead of its (near-)synonyms in medieval alliterative poems. However, the best poets are able to ensure that even in those cases form and meaning go hand in hand. The uses of ME skerren ‘to scare’ and ME rade ‘afraid’ in the corpus are good examples of the Gawain-poet's skill in this respect.

In his analysis of the portrayal of sodomy in the Sodom and Gomorrah episode of Cleanness, Frantzen (1996) discusses various inversions that the poet puts forward in order to highlight the moral depravity of the towns: what should be a heterosexual paradise (Cleanness, ll. 697–704) has been turned into a pitch-filled pit (Cleanness, ll. 1005–12) because what should be the practice of a despised minority has become the norm of the powerful majority. One further contrast, not explored by Frantzen, can be identified as the triggering factor for the use of ME skerren in this context: the portrayal of a mob of men who are not getting ready to fight an enemy in a manly way but are instead eager to engage in the cowardly act of killing a man for the sake of their desires (Cleanness, ll. 833–4). This behaviour implies that their unmanliness, specifically commented on in terms of their sexual practices (see Cleanness, l. 696), has wider implications for other areas of their lives. In this respect, ME skerren would have been chosen in l. 838 because of its usefulness as an alliterative partner for two key words in the development of the contrast with manly war, viz. ME scoute-wacch ‘sentinel, guard’ and ascri, a term that means mainly ‘battle-cry’.¹¹

In grete flokkez of folk þay fallen to his ȝatez;

As a scowte-wach scarred so þe asscry rysed;

With kene clobbez of þat clos þay clatz on þe wowez,

And wyth a schrylle scarp schout þay schewe þyse wordez (Cleanness, ll. 837–40)

(‘In great crowds of people they rush to the gates; the shout rose as though from a terrified watchman; with stout clubs they clatter on the walls of that enclosure, and with a shrill sharp shout they utter these words’)

In this well-populated lexico-semantic subfield, ME skerren was the only Middle English verb meaning ‘to feel fear’ and ‘to cause fear’ or related meanings that starts with the cluster /sk/ and this made it popular in alliterating contexts.¹² This formal usefulness might have contributed to its semantic broadening to refer to a state of perturbation, for a variety of reasons, be it fear, anxiety or displeasure (we see the fuzzy boundaries between emotions).¹³ That might be the meaning in Cleanness, l. 598, where ‘scarrez’ refers to God's reaction when he encounters repugnant actions:

He is so skoymos of þat skaþe, He scarrez bylyue;

He may not dryȝe to draw allyt, bot drepez in hast:

And þat watz schewed schortly by a scaþe onez. (Cleanness, ll. 598–60)

(‘He has such repugnance of that sin, He [scarrez] immediately; He cannot bear to hold back, but strikes quickly; and that was once suddenly shown by a disaster.’)

This context has caused some uncertainty amongst editors and lexicographers: while Andrew & Waldron (2007: 137 note to ll. 597–9, and 344, s.v. scarre) translate it as ‘is provoked’ and gloss the sense also as ‘react fiercely’ in the glossary to their edition, Putter & Stokes (2014: 126 note to l. 598, and 933, s.v. scarres) prefer ‘scatters, drives off’ as its meaning in this context. The OED (s.v. scare, v., sense 2) renders it as ‘to take fright, be scared’, while the MED (s.v. skerren, sense b.) takes advantage of the current polysemy of the verb alarm and associates this context with the meaning ‘to become alarmed’. Thorlac Turville-Petre (p.c.) prefers to understand the form as a reference to punishment (‘strikes out?’) and to identify it instead with a different verb, ME skairen, which also enjoyed some popularity amongst (alliterative) authors from Scandinavianized areas (see MED, s.v. skairen). If we are dealing with the same verb (ME skerren), these contexts might record different steps in the process of semantic broadening of the verb: (i) it is first attested in the Ormulum with seemingly strong negative connotations as it is only associated with the actions of the Devil, who enjoys terrifying humans already frightened by his presence instead of comforting them, as God's angels do (see Pons-Sanz 2015: 584–5; and MED, s.v. skerren); (ii) Cleanness, l. 838 records the verb meaning ‘to scare, frighten’ without any clear negative connotations; and (iii) in Cleanness, l. 598 the verb could be said to express a general state of perturbation, not necessarily or solely associated with fear.¹⁴ Yet it is important to be careful when presenting such a neat process of semantic change because the Ormulum is infamous for its repetitiveness and, therefore, it might portray a biased picture of the term that does not correspond with its actual (possibly wider) initial meaning.

Just as formal matters seem to have facilitated the only use of ME skerren in the corpus as a clear member of the fear subfield, they can also account for the two attestations of ME rade. This is most evidently the case as far as its only use in Cleanness is concerned, for there it helps to tie in the alliterative pattern of the line with its content through onomatopoeia (/r/ for the roaring of the frightened ox): ‘romyes as a rad ryth þat rorez for drede’ (Cleanness, l. 1543; ‘cries out like a frightened ox that roars for dread’). As in the case of ME skerren, the poet did not have many adjectives or past participles alliterating in /r/ at his disposal in this lexico-semantic subfield.

Formal matters are also likely to have dictated the choice of the adjective in its only other context in the corpus:

Þenn Arþour, bifore þe hiȝ dece, þat auenture byholdez

And rekenly hym reuerenced, for rad was he neuer,

And sayde: ‘Wyȝe, welcum iwys to þis place.

Þe hede of þis ostel, Arthour I hat;

Liȝt luflych adoun and lenge, I þe praye,

And quatso þy wylle is we schal wyt after.’ (SGGK, ll. 250–5)

(‘Then Arthur, before the high dais, beholds the strange happening and courteously greeted him, for he was not at all afraid, and said: “Sir, welcome indeed to this dwelling. I, the head of the house, am called Arthur. Kindly dismount and stay, I pray you, and whatever your wish is we shall learn afterwards.”’)

It might be the case that a pun is also intended here. The poet might have wanted his audience to bring to mind the homonymous native adjective and adverb ME rade, which refer to doing an action quickly, hastily or rashly. The narrator has just said that Arthur is somewhat childish and has a wild brain (SGGK, ll. 86 and 89); thus, the audience might be invited to interpret the statement sarcastically in relation to what is to come. After all, when faced with the Green Knight's disparaging words, Arthur is very rash in assuming that the visitor is looking for a physical fight even though the latter explicitly denies it (SGGK, ll. 276–8), in taking the ‘Christmas game’ upon himself (SGGK, ll. 323–5), and in establishing what he thinks is the best way to carry the challenge through without considering other options (SGGK, ll. 372–4).¹⁵ If this is the case, in this context we have an interesting combination of formal and semantic factors, which are further discussed in the next section.

3.2 Semantic nuances and the choice of terms for fear

On many occasions the Gawain-poet seems to move away from his preferred ME drēde and fēre word-fields because other terms allow him to explore the close relationship between fear and other emotions, as well as cognitive processes.

3.2.1 Fear and hatred (disgust)

The investigation of what moral righteousness involves, and the divine or human punishments that one can expect when it is not upheld, gives the poet plenty of opportunities to analyse the effects of the well-established overlap between fear and hatred, particularly disgust (HTE, 02.04.14.01.01), manifested as moral and/or physical revulsion (cf. Díaz Vera 2011: 92–3; Stefanowitsch 2006: 88–90). In some contexts it is harder for a modern audience to see the interaction between these emotions because of our limited knowledge of the semantic range of the terms involved; this is particularly the case with the use of ME skerren in Cleanness, l. 598 (discussed above, section 3.1) and the only attestation of ME grīen in the corpus, where it refers to Gawain's reaction when the Green Knight points out his flawed behaviour: ‘So agreued for greme he gryed withinne’ (SGGK, l. 2370; ‘so overcome with vexation that shuddered within’). This uncommon verb, of uncertain meaning and etymology, appears to refer to shuddering mainly as a response to fear (cf. ME grūen ‘to be terrified, shudder; to be troubled’; OE gryre ‘fear, terror, dread’; and SGGK, l. 2382, where we are told that Gawain has always been afraid to commit treachery and to break his troth) but also to distress more generally.¹⁶ The latter is important in the context under discussion, where Gawain's distress is caused by a combination of emotions, his shame and hatred/disgust for his recent behaviour being some of them (cf. SGGK, ll. 2372 and 2374).¹⁷ However, the overlap between the subfields of fear and disgust can be more clearly studied in relation to the uses of the loans ME uglī and hidour because we know more about the semantic ranges of these terms and their respective word-fields; consequently, the lexical choices around these terms are the main focus of this section.

The Norse-derived adjectives ME auelī and uglī could be said to be interchangeable because they both start in a vowel and refer to something that can provoke fear, but this interpretation would be missing important differences between them:

1. (i) ME auelī belongs to a much more central word-field in this lexico-semantic subfield than ME uglī. The centrality of the ME aue word-field in the Gawain-poet's idiolect and dialect is suggested by the number of derivatives in the poet's corpus (the adjective and adverb ME auelī, as well as auelēs, a term associated with courage: HTE , 02.04.22; see the Appendix) and the presence of the noun ME aue as the non-alliterative final stress in l. 234 of St Erkenwald.¹⁸ In fact, in these texts the Norse-derived words have completely taken over from their native equivalents (i.e. the ME eie < OE ēge word-field), which, again, shows one step further in the integration of this word-field into English than in the Ormulum, where the two word-fields still coexist (see Pons-Sanz 2015: 576, 583). That this might not be just a diachronic difference is suggested by the fact that the works by Robert Manning, a near contemporary of the Gawain-poet from south Lincolnshire, do record the ME eie word-field (see MED, s.v. eie).

2. (ii) From a semantic perspective, the two adjectives differ in that ME uglī can also refer to something loathsome, be it from a sensory or moral perspective (cf. ME uggen ‘to be fearful, fear; to feel loathing and disgust’). This polysemy seems to have arisen once the term was integrated into English, possibly by association with other terms that exhibit similar semantic overlap (e.g. ME (a)grīsen and the hidour word-field, on which see below); the Norse word-field (cp. Old Icelandic ugga, ugligr, etc.) referred only to fear and this appears to have been the original semantic association of the word-field in English (see MED, s.vv. uggen and uglī).

When the Green Knight first walks into Camelot, its inhabitants are scared by his imposing appearance because of his height and, obviously, his colour; he causes both fear and awe: ‘Þer hales in at þe halle dor an aghlich mayster, / On þe most on þe molde on mesure hyghe’ (SGGK, ll. 136–7; ‘there rushes in at the door a fearsome lord, the very biggest man on earth in height’).¹⁹ Later in the story, once his head has been cut off, we are told about his terrifying (ME uglī) body covered in blood and, more importantly, still moving and able to talk!

He brayde his bluk aboute,

Þat vgly bodi þat bledde.

Moni on of hym had doute,

Bi þat his resounz were redde. (SGGK, ll. 440–3)

(‘He twisted his trunk around, that fearsome body that bled. Many a one was frightened of him by the time he had finished speaking.’)

Bertilak's explanation in SGGK, ll. 2460–2 underlines that this disgusting trick was fully intended to terrify Camelot (see further section 3.2.3). While this episode is physically repulsive, the chapel where the Green Knight can be found is presented as a horrible place, frightening as well as morally loathsome. It is an inversion of the chapels Gawain is used to: it is not the well-kept place of prayer where one worships God but an unkempt place more suitable for the Devil to ‘pray’ in:

‘Wheþer þis be þe Grene Chapelle?

Here myȝt aboute mydnyȝt

Þe Dele his matynnes telle!’

‘Now iwysse,’ quoþ Wowayn, ‘wysty is here;

Þis oritore is vgly, with erbez ouergrowen[’] (SGGK, ll. 2186–90)

(‘“Is this the Green Chapel? The Devil might well recite his ‘matins’ here around midnight!” “Now, indeed,” said Gawain, “it is desolate here; this chapel is threatening, overgrown with plants.”’)

The use of ME uglī in Cleanness, where it refers to the forthcoming destruction of the morally corrupt towns, similarly brings together the association of something that is physically and morally detestable: ‘Bot þay wern wakned al wrank þat þer in won lenged, / Of on þe vglokest vnhap þat euer on erd suffred’ (Cleanness, ll. 891–2; ‘but those who lived in the town were woken quite awry by the most terrifying calamity ever suffered on earth’). The destruction of the towns must, of course, have been terrifying, but the event is presented as equally disgusting: not only were they engulfed by the ground because of their sins but, before this happened, there was a rain of fire and sulphur that brought about a foul smell (see Cleanness, ll. 953–5).

In this respect, it might seem surprising that the poet uses the adverb ME auelī and not uglī in a context referring to the threatening manner in which the Sodomites deliver their despicable message to Lot:

Þenne þe rebaudez so ronk rerd such a noyse

That aȝly hurled in his eres her harlotez speche:

‘Wost þou not wel þat þou wonez here a wyȝe strange,

An outcomlyng, a carle? We kylle of þyn heued!

Who joyned þe be jostyse oure japez to blame,

Þat com a boy to þis borȝ, þaȝ þou be burne ryche?’

Þus þay þrobled and þrong and þrwe vmbe his erez (Cleanness, ll. 873–9)

(‘Then the scoundrels so proud raised such a noise that their villains’ speech rushed menacingly into his ears: “Do you not know well that you live here an outsider, an immigrant, a knave? We will strike off your head! Who appointed you to be judge to disparage our pastimes, you who came to this city a churl, though you are a rich man?” Thus they jostled and pressed and crowded around his ears.’)

The presence of ME auelī could be explained by the fact that the adverbs ME uglilī and uglī are not recorded until the fifteenth century (see MED, s.vv. uglī, adv.; and uglilī). However, a more interesting explanation, in keeping with the themes of the narrative, can also be put forward. The key to understanding this lexical choice might lie again in the poet's desire to present the Sodomites as unmanly and morally corrupt, eager to take part in a battle for the wrong reasons and in a cowardly way. In this respect, it could be the case that the poet wanted his audience to associate the adverb, through phonetic similarity, with ME arghlī/arwelī ‘cowardly, disgracefully’ (see note 18). The poet exploits the implications of the cognitive metaphor argument is war, which involves the conceptualization of (angry) words as flying projectiles that people throw at each other in an antagonistic way.²⁰ The Vulgate here simply reads ‘vimque faciebant Loth vehementissime’ (Genesis 19.9; ‘and they pressed very violently upon Lot’), whereas in the poem we have menacing speech rushing or even striking into (‘hurled in’) Lot's ears. ME uglī might be the adjective that the poet prefers to refer to something frighteningly repulsive, but his choices can also be mediated by other factors.²¹

Like ME grīen, the French-derived ME hidour is only attested once in the Gawain-poet's texts (Patience, l. 367, quoted above as exemplification of timor servilis and below for ease of reference) but, unlike the verb, this noun and other terms belonging to its word-field (e.g. the adjective ME hidous) are widely attested in Middle English texts. They appear in contexts referring to fear or disgust (either because of something physical or because of the moral implications of an action), a double usage that goes back to their Anglo-French etyma (AF hidur and hidus, respectively; see Anglo-Norman Dictionary Online, hereafter AND, svv. hidur and hidus). In his only use of the term, the Gawain-poet brings the polysemy of ME hidour to the forefront in order to encapsulate a key message of the narrative. If we look at other contexts where we are given access to the feelings of those facing God's wrath and punishment for their actions, we simply see timor naturalis; they are scared to die, but there is no remorse for their actions and hence no forgiveness from God (e.g. Cleanness, ll. 385–90, 989–92). The people of Nineveh are similarly afraid for their lives but, very importantly, they also feel revulsion for their previous practices: ‘Such a hidor hem hent and a hatel drede / Þat al chaunged her chere and chylled at þe hert’ (Patience, ll. 367–8). The translation provided by Andrew & Waldron (‘such a terror seized them and a cruel dread, that their demeanour all changed and they grew cold at the heart’) does not capture the polysemy of the French loan, but one could argue that it is precisely the overlap of the two emotions that leads the Ninevites to change their ways, repent of their behaviour and hence be saved. A comparison between the poem and its biblical source, where fear does not feature (‘et crediderunt viri ninevitae in Deo et praedicaverunt ieiunium et vestiti sunt saccis a maiore usque ad minorem’, Jonah 3.5; ‘and the men of Nineveh believed in God: and they proclaimed a fast, and put on sackcloth from the greatest to the least’), helps to highlight the significance of its discourse in the poet's oeuvre.

3.2.2 Fear and mental suffering

The association of the lexico-semantic subfields of fear and suffering (HTE, 02.04.11) is also well rooted in the English vocabulary, as another example of the fuzziness that surrounds emotional boundaries. In terms of cognitive semantics, the two emotions share a metaphorical identification with an opponent/oppressor that causes physical or mental pain (see Kövecses 1990: 74–9; Fabiszak 2002; Díaz Vera 2011: 97–8). The Gawain-poet is also keen to stress the close relationship between these emotions. At times this is done by bringing together different terms referring to them: e.g. the presence of ME grēven (a term associated with suffering) and glopning (a polysemous term associated with fear, as well as suffering) in SGGK, ll. 2460–1 helps to highlight the emotional impact that the trick of a talking head detached from the body is intended to cause (see further section 3.2.3).²² At other times the poet relies primarily on polysemous terms that can refer to either emotion. This is particularly the case with two nouns: the French-derived ME affrai and the native ME cāre.

The ME affraien word-field is likely to have had connotations of physicality and violence because, besides being part of the lexico-semantic subfield of fear, these terms could also refer to launching a physical attack on someone.²³ This might have been the original meaning of the word-field if the generally accepted interpretation of the French terms as Germanic loans (< Latin *exfridare ‘to take sb. out of their peace’, cf. Proto-Germanic *friþu, OE frið, German Friede ‘peace’) is correct, with the ‘peace’ one is ‘taken out of’ referring, originally, to one's physical (and/or sociopolitical) situation rather than a mental state.²⁴ Such violent connotations are put to good use in Cleanness, l. 1553, which presents Belshazzar's terrified reaction when seeing a bodiless fist scratch letters on the wall. Here the experiencer is the ‘flesche’ rather than the ‘herte’, which further highlights the physical effects of his fear:

Sone so þe kynge for his care carping myȝt wynne,

He bede his burnes boȝ to þat were bok-lered,

To wayte þe wryt þat hit wolde, and wyter hym to say –

‘For al hit frayes my flesche, þe fyngres so grymme’. (Cleanness, ll. 1550–3)

(‘As soon as the king was able to speak again for his [care], he ordered his scholars to come, to examine the writing [to see] what it meant, and to tell him clearly – “for the fingers so grim utterly frighten my flesh”.’)

While the aphetic form ME fraien does not seem to have included suffering in its semantic range, ME affraien and its related noun, ME affrai, did, and we see this overlap of fear, suffering and the intensity of the emotions connoted by the word-field in the only other context where its members denote fear in the corpus:²⁵

My hede vpon þat hylle wattz layde

Þeras my perle to grounde strayd.

I raxled, and fel in gret affray,

And, sykyng, to myself I sayd:

‘Now al be to þat Pryncez paye’. (Pearl, ll. 1172–6)

(‘My head was laid upon the hill where my pearl slipped away in the earth. I stretched and fell into great [affray], and, sighing, said to myself: “Now may all be to that Prince's satisfaction”.’)

The polysemy of ME affrai helps to emphasize the strength of the emotional turmoil that the dreamer feels when he realizes that his beloved pearl has disappeared from his sight and is lost forever. In Cleanness, ll. 1550–3 it is mainly ME cāre, a term referring prototypically to suffering as well as general mental perturbation, rather than ME fraien, that highlights the complexity of the king's feelings (cf. Latin conturbare ‘to disturb’ in Daniel 5.6 and 5.9).²⁶

A medieval audience is likely to have been familiar with the effect that the various meanings of these nouns bring to these contexts, but capturing them in modern translations is difficult. This has contributed to some scholarly disagreement about how best to interpret ME affrai and cāre in these contexts: e.g. while MED (s.vv. cāre, sense 4.a; and affrai, sense 2) seems to associate them with fear, Andrew & Waldron (2007: 174, note to ll. 1550–2; and 108–9, note to ll. 1171ff.) prefer to analyse them as references to suffering instead. However, breaking these semantic connections apart detracts from the overall effect of the terms in their attempt to capture the dreamer's and the ruler's emotional disturbance. Anxiety, which can also refer to the two emotions in Present-day English, might be a good translation for the two nouns. This is precisely the term that Putter & Stokes (2014: 826, s.v. care) choose to render ME cāre in the other context where the overlap between the two emotions is brought to the forefront. As we have seen in section 2, while the Green Knight rather generously attributes Gawain's fault (not honouring in full the exchange of winnings agreement) to his love for life, Gawain blames it on his anxiety about the blow that he was expecting to receive: ‘For care of þy knokke, cowardyse me taȝt / To acorde me with couetyse, my kynde to forsake’ (SGGK, ll. 2379–80; ‘because of [care] about your blow, cowardice taught me to reconcile myself with covetousness, and to forsake my nature’). Here ME cāre makes us think about his fear as well as the mental distress (‘þro þoȝtes’; oppressive thoughts) that he experiences in his dreams before the Lady of Hautdesert comes to his room on the morning when he accepts the girdle (SGGK, ll. 1750–4), an acceptance that is also presented in terms of patient suffering and tolerance (SGGK, l. 1959).

Interestingly, other than his reference to cowardice in the long list of moral failings that he attributes to himself, neither Gawain nor the narrator ever acknowledges openly that he felt fear during the challenge, this use of ME cāre being the closest he comes to it. This can be associated with the poet's attempt to bring together two seemingly irreconcilable demands on chivalric emotional communities: to face death bravely and, at the same time, to fear it appropriately. In the text, knights strongly deny that they feel fear (see SGGK, ll. 2210–1) with the half-hearted help of the narrator.²⁷ Characters and narrator manage instead to achieve a difficult compromise by associating these two approaches to fear with gender, as noted by Yeo (2016). The only two uses of the French-derived ME dismaien in the text (and the whole corpus) present an example of this approach: while we are told that the Green Knight, the epitome of virility, is not at all ‘dismayd’ (l. 336; nor ‘bayst’, l. 376) by the fact that Gawain has taken on his challenge, Arthur encourages Guinevere, as representative of all the ladies in Camelot, not to ‘demay [herself] neuer’ (l. 470) because of what has happened in front of her, which implies that she, as a woman, is likely to feel that way.

3.2.3 Fear and uncertainty

Medieval audiences were familiar with the fact that fear and the emotional distress it is associated with are closely linked to one's cognitive processes, to the extent that they can hinder or stop orderly thought. Thus, Aquinas explains that ‘if fear should develop to the point where it upsets reason, then even mental functioning will be disturbed’ (Summa theologiae, I–II, 44.4). Our poet also touches on the interaction between mental capacity (HTE, 2.1) and emotion through his choice of terms for fear. While the French-derived ME dismaien and baishen seem to have partially referred to a disconcerting feeling provoked by fear, as native terms did before them (e.g. OE aþracian; see MED, s.vv. baishen and dismaien; DOE, s.v. aþracian; and section 3.2.2), the French-derived ME doute word-field enables the poet to highlight instead a feeling of uncertainty or perplexity that can lead to fear. That is the case in all the uses of the word-field in the corpus bar one (viz. SGGK, l. 784); they are recorded in SGGK and are associated with the challenge that the Green Knight proposes. While in SGGK, l. 222 we are told that the Green Knight does not fear the possible events that he might face in Camelot (‘Driuande to þe heȝe dece – dut he no woþe’; ‘making for the high dais – he feared no danger), l. 246 refers to the fear felt amongst Arthur's knights, which arises from being faced with a green visitor whose attire, actions and words send contrasting messages (cf. ME auelī; see section 3.2.1):

As al were slypped vpon slepe so slaked hor lotez

In hyȝe –

I deme hit not al for doute,

Bot sum for cortaysye – (SGGK, ll. 244–7)

(‘Their voices died away as though they had all fallen asleep suddenly – I judge it not wholly for fear but partly for courtesy’)

Once Gawain has cut the Green Knight's head off, fear arises in l. 442 (quoted in section 3.2.1) from the unnatural fact that a bodiless head can continue to talk. As noted above, Bertilak pays particular attention to this trick in his final explanation about the point of the adventure; he highlights that the trick of the talking, bodiless head was intended to cause deep astonishment and terror:

Ho wayned me þis wonder your wyttez to reue,

For to haf greued Gaynour and gart hir to dyȝe

With glopnyng of þat ilke gome þat gostlych speked

With his hede in his honde bifore þe hyȝe table. (SGGK, ll. 2459–62)

(‘She sent this marvel to deprive you of your senses, in order to distress Guinevere and cause her to die from terror at that man who spoke in supernatural manner with the head in his hand before the high table.’)

Cognitive and affective processes (‘your wyttez to reue’, ‘greued’ and ‘glopnyng’) are brought together to draw full attention to the amazement, fear and general distress that Morgan le Fay was hoping to cause to the inhabitants of Camelot in general, and to Guinevere in particular.

The high concentration of occurrences of the doute word-field in connection with the initial interaction between Camelot and the Green Knight emphasizes the astonishing nature of the events that unfold during the Christmas celebrations. The other significant occurrence of a member of the word-field comes at the end of the romance, when Gawain is about to meet the Green Knight again. Not knowing how he will fare when he receives the blow that he is owed is what terrifies him, even though he is determined to pretend that is not the case and to control his body even if he cannot placate his mind:

He lened with þe nek and lutte

And schewed þat schyre al bare,

And lette as he noȝt dutte;

For drede he wolde not dare. (SGGK, ll. 2255–8)

(‘He lent and bowed his neck and showed the white flesh all bare, and behaved as though he feared nothing; he did not intend to flinch for fear.’)

ME douten puts Gawain in direct contrast to the Green Knight. The latter, who represents otherworldly manliness, enters the unknown with the comfort of knowing that he has come to a Christian court renowned for its good manners and courtesy, and that he has the protection of an accomplished sorceress. In contrast, Gawain, an example of human fallibility, goes to face his fate at the hands of a man described as ‘methless and mercy non uses’ (l. 2106; ‘intemperate and practises no mercy’) with the help of a magical object of untested powers.

4 Conclusions

This article has explored the make-up of the lexico-semantic subfield of fear in four late Middle English texts renowned for their lexical richness and, in doing so, it has helped to redress the imbalance between the significant attention that this emotion and its expression in Middle English texts have received from literary critics and historians, and its limited treatment by historical linguists. Written under the auspices of the Gersum Project, the main aim of the article was to exemplify the benefits of the in-depth study of the Norse-derived terms on the basis of their form and its contribution to the alliterative structure of the texts (stylistics), their semantic range (semasiology) and the other (near-)synonyms in the poet's idiolect (onomasiology) in order to gain a better understanding of their process of integration into and accommodation to medieval English. In spite of the significance of such understanding, this is an aspect in the study of the lexical effects of Anglo-Scandinavian interactions that has received less attention than the identification of the Norse loans themselves.

The integration of Norse-derived terms into English was a complex process, with different issues at play. This article has shown that the form of some Norse loans (ME skerren and rade; see section 3.1), in particular the nature of their onset, was an important factor for their use in contexts where alliteration played a significant role because they contributed to the expansion of the range of alliterating environments the lexico-semantic subfield could participate in. In some cases, the existence of native cognates which were central to the lexico-semantic subfield can also be seen to have facilitated the integration of the Norse loans and the expansion of their word-field into English (cf. the ME aue and eie < OE ēge word-fields; see sections 2 and 3.2). Their meaning and the nuances that they could help express have similarly been shown to have greatly contributed to their integration. Consider, for instance, the semantic differences between ME auelī and uglī; the latter shared with the French-derived ME hidour word-field the ability to bring together the fear and hatred (particularly disgust) lexico-semantic subfields. This article has explained how the Gawain-poet relied on such semantic matters for clear thematic purposes and has thus exemplified the need to consider the whole semantic range of a term when studying its use in a particular context.

Because of the necessity to explore the interaction between the Norse-derived terms and the other members of the fear subfield, the significance of this article goes beyond the boundaries of the lexical effects of Anglo-Scandinavian interaction. The article's second aim was to analyse the impact that language contact more broadly had on the expression of fear in Middle English. The discussion in section 2 has shown that the Gawain-poet relied heavily on native terms as the core members of this lexico-semantic subfield (the ME drēdren and fēren word-fields). However, he was also keen to exploit the polysemy of his vocabulary in order to enrich the message being transmitted, regardless of whether the terms that allowed him to do so were native (e.g. ME cāre) or had originally been borrowed from either Old Norse or French (e.g. the French-derived ME affraien and douten word-fields and, possibly, glopning, a term that might be Norse-derived). In this respect, this article has led the way in engaging with recent scholarly calls to explore the impact that medieval multilingualism had on ‘the range of emotional utterance in Middle English texts’ (see section 2), bringing together two important areas of research where there is still much more work to be done, separately and in conjunction.

Indeed, understanding how loans helped speakers to talk and think about particular emotions can be of mutual benefit to both areas of research, as this article has demonstrated. On the one hand, in terms of research into the expression and conceptualization of emotions, this article has shown that with this approach we can better understand processes of semantic change within a subfield (e.g. the use of ME arghen with the meaning ‘to inspire fear or awe’; see above, note 18) and the shaping of cognitive associations. We have seen that the overlap between the expression of fear and other emotions, such as disgust (section 3.2.1) and suffering (section 3.2.2), and between fear and mental capacity (section 3.2.3) was already well established in Old English. Loans (e.g. the French-derived ME hidour word-field) helped to maintain these associations, which in turn facilitated the semantic change of some borrowed word-fields once integrated into English (e.g. the Norse-derived ME uglī word-field and, possibly, skerren) and thus further strengthened these associations. At the same time, loans also allowed the development of further nuances within such well-established interactions (e.g. the French-derived ME doute word-field). On the other hand, with regard to the study of the impact of multilingualism on medieval lexis, through its focus on a lexico-semantic (sub)field conceptually very far from the technical fields that French influence is commonly associated with, this article has re-emphasized the need to rethink the sociolinguistic situation of medieval England in order to present a more nuanced account of the perception, use and influence of the various languages spoken at various times across the realm.

Finally, even a superficial comparison between the manifestations of the lexico-semantic subfield of fear in the Gawain-poet's fourteenth-century texts and the twelfth-century Ormulum has shown interesting results in relation to the integration of the Norse-derived terms into Middle English, in terms of their possible semantic development (ME skerren) and their competition with native (near-)synonyms (ME aue and eie word-fields). A detailed study of the make-up of particular lexico-semantic fields, based on texts from different periods and dialectal areas, still remains a key desideratum for the study of the integration of Norse loans (as well as loans from other languages) into medieval English. Bringing together semantic, stylistic, diachronic, diatopic, and, whenever possible, diaphasic and diastractic matters would take further McIntosh's (1989) message that historical linguists need to engage in Middle English word geography in order to advance our knowledge of medieval vocabulary in general, and the significance of Norse influence in particular.