Rereading The Wife’s Lament with Dido of Carthage: The Husband and the Herheard

Abstract

The Old English poem in The Exeter Book titled The Wife’s Lament is about longing and loneliness; the woman speaking in the poem longs for her absent husband who has sent her to live in a “cave under an oak tree”. The husband’s attitude toward his wife is a major point of controversy among commentators on the poem: has he sent her there as a punishment or for her protection? This essay argues that he loves her and seeks to protect her in his absence. The argument supporting this view addresses the following three topics: the reason he must leave and his brooding silence preceding that departure, the culture of warrior oaths, and the nature of the “cave” where the speaker is located. The first two discussions assess and reframe previous scholarship, while the discussion of the speaker’s location introduces a new area of research, the archaeology of early medieval rock-cut buildings. Finding that the poet might imagine the Wife inhabiting such a constructed building invites us to think about her, her husband, the poem, and even the Exeter Book itself within a new and interesting real-world context.

1. Introduction

In 2003, John D. Niles published an article titled “The Problem of the Ending of The Wife’s Lament” (Niles 2003) that he then revised and incorporated as Chapter 5 in his book The Old English Enigmatic Poems and the Play of the Texts (Niles 2006). The Wife’s Lament1 is a poem in the voice of a young woman2 living in an isolated location “under an oak tree”3 after the departure of her felaleof (much-beloved) husband,4 whom she misses desperately. Variations on the word for “longing” occur at key points throughout the poem (lines 14, 29, 41, 53: see Stevick 1960, p. 21). The issue that Niles addresses is “how to construe a passage of ten and a half lines at the poem’s close (lines 42–52a), whether as the speaker’s gnomic reflection on the sorrows of life or as her outright curse upon a man who has wronged her” (Niles 2006, p. 149). He explains that “by combining the methods of philology and historical anthropology, I hope to present an interpretation of The Wife’s Lament that is both linguistically sound and historically plausible” (Niles 2006, p. 150). Niles emphasizes that interpreting lines 42–52a as the Wife’s curse “upon a man who has wronged her” is only one way of reading those ambiguous lines.5 While agreeing that other readings of those lines have merit, this essay finds value in Niles’ argument for the Wife’s curse but stands firm against his borrowed idea of a cruel husband who has wronged her. To support the idea that the Wife curses her husband, Niles provides a long and fascinating sidebar essay on “cursing as a social institution and a literary theme, both in the Middle Ages in Europe and, as space permits, in other times and places” (Niles 2006, p. 150).6 But space is, as he indicates there, an issue in his lengthy chapter, so to get quickly to his discussion of cursing and avoid the distraction of problems earlier in the poem, he calls on an acknowledged expert on The Exeter Book “elegies”, Anne Klinck.7 In her otherwise exemplary book, Klinck interprets the husband as cruel and even wishing harm to his wife—a good reason for her to curse him—but the text does not stand up well to that reading, as will be shown. Notably, reading the husband as cruel misses the similarity of the Wife’s situation to that of the most powerfully cursing woman in all literature, Dido of Carthage, whom Aeneas sadly abandons under pressure from the gods. (Dido’s curses against him have such force that, according to Virgil, they echo down the generations to incite the devastating Punic Wars of 264 to 146 BC.) Dido will be our ally in amending Niles’ argument to allow for the idea of a husband who, like Aeneas,8 cares about the wife he is forced to leave while arousing her fury by doing so and causing her to curse him in a moment of anguish.

What follows is organized into three sections:9 First comes a reading of the poem that does not make the man an enemy to the Wife; this is followed by a sidebar essay on oath-taking that supports the thesis, and the essay concludes with a discussion that examines the Wife’s “cave under an oak tree” as a possible rock-cut building. While rock-cut temples and even entire cities are known tourist sites in other countries,10 the more modest English rock-cut buildings, typically dwellings cut into a cliff-face, draw little attention and have thus been overlooked as a source for the poet’s possible model for the Wife’s “cave”. Each of these three sections reaffirms the central thesis that the husband may genuinely care about his wife as her “friend” (lines 25, 47, 50).

Accepting Niles’ reading of lines 42–52a as the Wife’s curse, while not interpreting her husband as hostile toward her, reveals the passage to be dynamic self-narration. Like Dido, the Wife is perfectly capable of cursing a felaleof (much-beloved) husband who, due to a conflicted situation, leaves her in lonely misery (without friends: lines 10, 16–17, and 33–36). Central to the argument for a caring husband is a positive meaning for the word or words in the letters herheard in line 15 (Hēt mec hlāford mīn herheard niman), understood in five different ways by a series of respected scholars, with negative interpretations by Klinck and some others. I argue, along with Karl P. Wentersdorf (1981) and P. R. Orton (1989), that the most reasonable way to understand herheard is to interpret it as the compound herh-eard (compare modern “churchyard”), referring to the site of a herh or “pagan temple” (hearg in West Saxon),11 where the husband has sent the speaker of the poem for her safety.12 In Enigmatic Poems, Niles confesses to finding “Orton’s reading persuasive”, but since “it is adopted by neither Klinck nor Muir” he chose not to rely on it in his essay (Niles 2006, p. 152, n. 7). Instead, he adopted Klinck’s scenario of the hostile husband.13 But Niles’ reading of the poem as “an imagined cri de coeur” by a woman cursing her husband because she longs for him (Niles 2006, p. 206) works even better for his argument when the husband is seen as entangled in a web of competing obligations—as Aeneas was. With the cruel husband removed, it is possible to draw attention to the Wife’s apparent understanding, even approval, of the man’s choice of action, the only honorable choice within the constraints of his culture (Notably, Virgil describes Aeneas as pius, right-thinking, when he makes a similar choice; Aeneid 4: 393). This recognition of the husband’s choice as culturally correct complicates the speaker’s emotional response to his departure, revealing a woman who is socially adept and aware of the mores of her culture, even while furious, and a poem that is more nuanced than previously thought. There is reason to suppose that the Wife would understand what her husband must do, and even why it is appropriate for him to be silent about his decision to take violent action. That violence is not the source of her complaint, nor does she ever complain about the physical nature of the cavern to which he has sent her. She calls it “old” but does not describe it as damp or chilly; the surrounding deep dales and too-high hills are what she seems to find oppressive. As she says at the very beginning of her lament, the cause of her sufferings, “ever since I grew up” (line 3b), is the near-total isolation in which she finds herself.

In saying siþþan ic up weox, a phrase usually overlooked or misunderstood (one writer taking it to mean “throughout her life”), the Wife tells us something about herself that may help to clarify her present situation. When she speaks about “what sufferings I’ve experienced ever since I grew up,/both new ones and old, never more than now” (lines 3–4) and then immediately tells about her husband’s departure (lines 6–8), this may suggest that, once she had reached a suitable age (line 3b; often very young), she was married exogamously, almost certainly for political reasons, as was expected for a young woman of elite status. Thus, one can assume that she was a child-bride (in our terms) wrenched from family and friends, and now, with her husband also gone, her “friendless” situation is even worse. Leonard Neidorf makes a good case for the similarity of her situation to that of Hildeburh in the Finnsburh episode in Beowulf, describing that lady as “a Danish princess living in Frisia, possibly against her will, who perhaps had few friends to begin with and has now lost the people closest to her” (Neidorf 2017, p. 199). Like the Wife, Hildeburh experienced sufferings both “old” when taken from her Danish birth home and “new” upon losing her loved ones in Frisia, in the now of the Finnsburh story. A significant difference lies between the two women’s fates, however: Hildeburh’s son and brother have been slain, whereas the Wife’s husband has merely gone over the sea. After Hildeburh receives satisfaction (as Neidorf persuasively argues), she is escorted back home to Denmark, but the Wife will remain in her cave in a foreign land, waiting (line 53).

In her introduction to The Wife’s Lament, Anne Klinck says, “My understanding of the relations between man and wife hinges on reading her heard (line 15), where I take heard as the adjective “cruel” applied to the husband, and by translating mines felaleofan fæhđe (line 26) as ‘the enmity of my very dear one’” (Klinck 1992, p. 50). In this section of the essay I will first examine those two words herheard and fæhđu, reading them, as Peter S. Baker does, in their most obvious meanings14 to absolve the husband of hostile intentions toward the speaker. Then I will consider how the verbs begietan (used twice) and hatan (het) may further clarify relationships in this poem. Also relevant is the punctuation after gebæro at line 21, another important word appearing twice, with its ethical nuance glossed at line 44.

Attention to herheard comes first because so much of the poem’s meaning rests on the nature of the Wife’s location and her attitude toward it, or, more accurately, her attitude toward being left there alone. The look of the word in the manuscript creates this crux in the poem’s meaning. Even though the line-end division of a word is arbitrary in The Exeter Book, those who argue for a cruel husband make much of the fact that the eight letters of herheard are divided by the end of a manuscript line to read her heard, and they interpret this as what the word-division makes it look like: the adverb her (here) followed by the adjective heard (hard, interpreted as cruel15), used nominally. Read this way, line 15 produces a meaning like “here [where the speaker is located] the cruel man commanded me to make a home”. But Klinck (following the arguments of two previous scholars16) makes the pronoun “me” the object of the verb niman (take or hold) instead of het (commanded), so that she can then read line 15 Hēt mec hlāford mīn her heard niman and translate it as “My lord commanded, cruel [hard], to seize me here” (Klinck 1992, p. 181). Not only does the verb “seize” exaggerate the usual meaning of niman (“take”), this translation also produces a syntax as awkward in Old English as in modern English when it leaves the transitive verb het (“commanded”) without a direct object: whom did he command to seize her? Most editors of the poem interpret the line in a straightforward way. Recognizing that herheard niman in line 15b echoes the idiomatic phrase eard niman, “to make a home” (i.e., settle in a place), R.F. Leslie interprets herheard as two separate words and emends it to her eard, “here, [an] abode”, indicating the place where the Wife has been told to take up residence. But Leslie’s word division creates a metrical problem because adverbs like her do not normally alliterate, and eard, now the main stressed word in the second half-line, does not alliterate. Thus Leslie’s line Het mec hlaford min her eard niman does not scan as well as it might.17 R.D. Fulk’s corrective solution, to take herheard as meaning hirad (household, family), is clever but unnecessary and requires special pleading, as well as emendation.18 Both Wentersdorf and Orton, like C.W.M. Grein before them, prefer to retain herheard as an intact word, resulting in a simple declarative sentence: “my lord commanded me to dwell in this temple-yard” (compare the words churchyard and graveyard), or “to take a grove-dwelling” according to Orton (1989, pp. 209–10). Wentersdorf is further convincing with his detailed argument for interpreting herheard as a temple that would traditionally be located in a grove, as it is in this poem at line 27b: on wuda bearwe. Which among these opposing readings is preferable depends on the reader’s understanding of the husband’s intentions toward his wife, whether he wants her out of the way as a nuisance, or even killed, as Elinor Lench suggests (Lench 1970, p. 16), or means to send her to a place where she will be safe, perhaps safe from his family. There is no doubt about his family’s negative attitude toward the woman because she states it clearly: they are secretly trying to separate the couple as far apart as possible (lines 11–13). Klinck defends her argument for taking heard to mean “hard”, indicating a cruel husband, by claiming that the spelling of herh for West Saxon hearg (“pagan temple”) is too unusual to be a valid option (Klinck 1992, p. 180) and that therefore herheard is an “unlikely form” (Klinck 1992, p. 50). The very existence of a word herheard is “dubious”, she concludes (Klinck 1992, p. 181). As Peter Baker points out, however, herh occurs at line 3072 of Beowulf (hergum geheađerod, “confined in a pagan shrine” (Peter S. Baker 2013, p. 208, n. 7)), and approximately the same spelling arguably appears on the Franks Casket in the compound herh-os (“temple deity”).19 If the word had not been misdivided (according to modern usage norms) where it occurred at the line-ending, herheard would probably have been universally recognized to be the compound herh-eard, as Grein, Wentersdorf and Orton understood it, and an unbiased reader would accept this word as a compound. There is no good reason to complicate this issue. Line 15, Het mec hlaford min herheard niman, should be translated as simply as it appears in the text with het, hlaford and herheard carrying the alliteration: “My lord commanded me to reside in [this] herheard”. The Wife does not appear to be disturbed by the location’s pagan associations, but that may be because the poet imagines her, as Niles suggests, in a “raw and primitive past” (Niles 2006, p. 206).20 The herheard is simply a “safe house” for this endangered young woman, but it is also a cave that she detests because of its loneliness.21

The next word to be addressed in this network of cruxes is fæhđu at line 26, commonly translated “feud”, and the controversy concerns what the Wife means when she says that she is suffering because of mines felaleofan fæhdu (“the feud[?] of my much-beloved”). Klinck translates fæhdu “hostility” (Klinck 1992, p. 183), a meaning well-established in corrective essays by G. Halsall (1999) and J. D. Niles (2015), but she takes this to indicate that the man’s hostility is directed against the speaker as he “cruelly commands his wife to be seized” (Klinck 1992, p. 51). Leslie understands fæhđu to mean “feud” in the sense of a recurrent hostility between groups, not between individual persons, and he claims that the word “cannot, therefore, refer to hostility of the husband toward his wife” (Leslie 1961, p. 55). Even when translating fæhđu as “hostility”, the more general meaning now assigned to that word, mines felaleofan fæhđu, may be understood within the context of the poem as the husband’s hostility directed against an unspecified enemy not within his kin group. The obligation to perform the dangerous act of violence about which he has been secretly thinking (morþor hycgendne, line 20) is what requires him to travel elsewhere, possibly to carry out an oath-duty to family or sovereign. It is the consequences of his hostility toward a third party, resulting in his departure on a mission that takes him from her, that has caused the Wife to suffer from his hostility (fæhđu dreogan) by being banished to her present isolation.22

Up to this point in the poem, the Wife resents the loneliness caused by the circumstance of her husband’s departure; she is not yet angry at him personally. This reading is supported by what she says later when she speaks of herself walking about alone in a “joyless lodging” (line 31) while she imagines lovers lying companionably together in bed (lines 33–34). Her situation in a land foreign to her23 was compared to that of Hildeburh above, but it may also be similar to that of another princess in Beowulf, the Scylding Freawaru, when her father marries her off to Ingeld the Heathobard in a bid to keep the peace between potentially hostile tribes (Beowulf lines 2026–29a). Beowulf imagines an angry old Heathobard warrior inciting Ingeld (lines 2041–56) until that prince’s love for Freawaru “becomes cooler” (lines 2065–66), so that the hostility between the Heathobards and Scyldings flares up again. Once the adjective heard (hard, cruel) has vanished into the noun herheard, there is no textual evidence of a similar spousal chill in The Wife’s Lament. The Wife is sad that the vows never to part that she and her husband took together are broken, “as if their friendship never were” (lines 24–25a), but if a stronger obligation takes her husband away, this is not an “actionable” breaking of those vows by a husband who “fails in his duty” (Niles 2006, p. 196). In most situations, obligation bound by oath to a strong corporate entity takes precedence over that to a private individual (See the sidebar essay for evidence that loyalty oaths made in earlier times might provoke a kind of compulsion verging on obsession).

How, then, does the wife herself interpret her husband’s command to take up residence in the herheard cave? Does she feel exiled against her will like a criminal (as some readers argue), or would she have understood that he wanted her there for her own good? The verb hatan in the phrase het mec hlaford min (line 15) can be read either way, as also could the adjective heard, if one chose to take her heard as adverb plus noun. If the husband is concerned about his wife’s safety when he is not there to protect her, he might tell her quite firmly to take shelter, which she could interpret as being “hard” on her. Even exaggerating the phrasing in this way, however, there is nothing to suggest that the Wife imagines her beloved freond as her enemy. But she does get very angry at him.

Niles supports his cruel husband reading with the verb begietan (begeat in lines 32 and 41). His idea, following Klinck, is that the Wife is seized (begeat) by enemies (Niles 2006, p. 156), rather than, as I and others interpret the situation, by the anxiety caused by her husband’s departure (fromsiþ frean, line 43). This interpretation of begeat may be fostered by a difference in our way of expressing emotional episodes. Whereas in modern English we would use the passive voice (“I was seized by anxiety”), in Old English anxiety actively grips the Wife. This happens to her repeatedly and wraþe (“fiercely”) in response to her isolated location; as Leslie says, in line 32, “The key words are oft and her” (Leslie 1961, p. 4). Thus, the sentence Ful oft mec her wraþe begeat fromsiþ frean (lines 32b-33a: “Very often, here, my lord’s departure fiercely seized me”) apparently refers to sudden involuntary grief spasms, a well-known physical reaction to overwhelming sorrow, especially that caused by the absence or death of someone greatly loved. It may seem to the sufferer as though they are taken over, fiercely seized, by an uncontrollable outside force.24 Similar metaphorical usage where an emotion is reified as a living antagonist occurs in The Wanderer as the protagonist experiences sorrow as a cruel companion (line 30) who binds him when aided by sleep (lines 39–40), but only the speaker of The Wife’s Lament seems subject to the severe spasms caused by panic about being abandoned.25

However, why does her husband have to leave her? Why not tell his anxious and mournful wife where he is going? Why must he hide his plans for violence under a blithe gebæro (“blithe bearing”), a feigned expression of ease? The description of his actions combined with her lament about the broken oath suggests that he may be under an oath that takes precedence over the oath they made together. Following Niles’ example in exploring themes in The Exeter Book that resonate intertextually between poems (Niles 2019, p. 72), let us return to The Wanderer where a central but rarely remarked preoccupation in that most famous of the elegies can help to answer these questions about The Wife’s Lament, and to understand the constraints that may be affecting the husband as well. The Wanderer poet is so deeply concerned about oaths made or carried out without sufficient preparation that it is surprising that no one (to my knowledge) has argued that this is a major theme in the poem, almost defining its secular level. Closely related to this theme of over-eager oath-taking is that of stoic (masculine) silence under pressure, what Robert E. Bjork, one of the few who have studied it, calls “the imagery of silence” in The Wanderer (Bjork 2008). This dual preoccupation in The Wife’s Lament with oath-taking (doing things with words) and silence (suppressing words) bears directly, I believe, on the husband’s outward expression of calm when seething inside with thoughts of mordor (violence, probably lethal; line 20). It may be presumed that he has departed “over the waves” (line 7) to enact that violence.26 There is controversy about how to punctuate the descriptive words bliđe gebæro at line 21, depending on whether that phrase belongs with what precedes or with what follows. The gnomic phrasing of lines 42–44, declaring that a man sad or serious of mind having “hard” thoughts should maintain a blithe bearing, suggests strongly that bliđe gebæro belongs with what precedes. This is an issue of some importance because it reflects the poet’s attitude toward the husband. A period placed after bliđe gebæro at line 21 reveals the husband, hygegeomor (sad in thought), to be silently contemplating violence under a “blithe demeanor”. When the poet uses that same phrase along with other echoing language in the later lines, he places positive value on behavior exactly like that of the husband in the earlier passage: a man ought to contain his anger silently under an easy demeanor while he is sad and suffering with long-lasting sorrows (geomormod, line 42, sinsorgna gedreag, line 45). The Wife, even while resenting his leaving despite the vows they had made together never to part, would be aware that her silent, abstracted husband has been thinking about a dangerous undertaking involving violence, and she would probably also be aware that the culture demanded that he should control his feelings silently under outward calm until he has devised a plan because “a man must wait”, as The Wanderer poet declares, “until he understands clearly” what action to take.

2. Speaking Big Words: Toward an Exploration of Problematic Oath-Taking

Beorn sceal gebidan, þonne he beot spriceð,

oþþæt collenferð cunne gearwe

hwider hreþra gehygd hweorfan wille. (The Wanderer, lines 70–72)

A man (beorn) must wait, when [i.e., after] he speaks an oath,

until, keen-spirited, he understands clearly

whither the thought of his breast will turn.

• Treow sceal on eorle. (Cotton Maxims, line 32b)
  Good faith must be in a man.
• Til biþ se þe his treowe gehealdaþ. (The Wanderer, line 112a)27
  Good is he who holds to his troth (ambivalent: he who is faithful to his vow or he who holds to his religious faith).
• Ec veit ein at aldrei deyr/dómr um dauđan hvern. (The Old Icelandic Hávamál, stanza 77: 3–4).

I know one thing that never dies,/a dead man’s reputation (dómr).

The multivalent term “oath” can refer to three different verbal acts, an assertion of truth referring to a past fact (as when witnessing in court), a promise of a future action (like an IOU), and a curse making use of a word once associated with the supernatural, usually a free-floating expletive (swear word).28 This essay is concerned only with the second of these, the promissory oath, and this comes in two forms: the “heroic oath” that promises performance of a designated valorous act in the near future, and the more open-ended and trusting “loyalty oath” that promises to perform an undetermined act whenever it shall be mandated. A promissory oath of either kind implicitly contains two elements, the “big words” making a commitment now to a later action, and the obligation to later perform the possibly dangerous deed that was sworn to or will be commanded. Thus, an oath (like any promise) is a double-tensed indexical (an indexical is a word or phrase that changes meaning with a change of context29). The two tenses implied by an oath refer to the time at which the words are spoken and a projected later time when that commitment must be honored in action.30 The concept of “honor” is crucial to oath-taking because any formal oath, whether heroic oath or loyalty oath, assumes that the oath-maker’s honor as a truth-teller is an important element of their perceived identity and that this factor will constrain them to “honor” their vow. Moreover, when a warrior is speaking an oath he challenges a future that he may hope for but cannot know, and the ritual sanctification of the oath within a male bonding situation engages the supernatural in the maker’s promise (as when the witness in a courtroom lays a hand on the Bible). All these factors constrain the oath-taker from choosing a future path in conflict with the intention of the oath, and this limitation on his freedom is relevant to evaluating the husband’s actions in The Wife’s Lament.

While attention to the philosophy of “promising” is currently vibrant with discussions extending into numerous disciplines,31 it does not, apparently, extend into the study of the ethics of warriors’ oath-taking implicit in several Old English poems, especially those evoking an imagined heroic age. The oaths and vows in both heroic and elegiac poetry are often at the heart of a situation that turns upon the issue of whether the promise is honored or broken. The main point of the brief study that follows is that there are degrees of obligation in oaths spoken publicly. Whether it is a vow to perform a specific deed or a declaration of open-ended fealty to a leader, when a promissory oath is sealed by the mead cup before the assembled warband, it will take precedence over a vow made to a private person outside that group, even to a wife.32 Proclaiming either kind of oath was and is a socially bonding act, but the performance of an oath in the meadhall before cheering witnesses must have been a special rite of male bonding. The exhilaration of a situation when spirits were high and the mead was flowing would have encouraged warriors, especially eager young ones, to vow to perform deeds of increasing danger and valor, thus eliciting the “heroic” oath. The dark side would come the next morning with the sobering realization that a witnessed oath must be carried out lest honor be lost. As everyone who has read Beowulf knows, the best thing a warrior in this (secular) world can hope for is the dom that lives after him, the good opinion of the group that will “deem”, or judge and assign glory, and that dom for a warrior is won by the honoring of an oath in action:

Ure æghwylc sceal ende gebidan

worolde lifes; wyrce se þe mote

domes ær deaþe; þæt biđ drihtgum[an]

unlifgendum æfter selest. (Beowulf lines 1386–9)

Each of us can expect an end

of life in the world; let him achieve who can

dom before death; that for the nobleman

no longer living is afterwards best.

The dom or good reputation (glory) that Beowulf speaks of in this passage is earned by lofdædas, and lof is the storied record of deeds performed well: “We have heard”, says the Beowulf poet, hu đa aþelingas ellen fremedon (“how the athelings performed acts of courage”, line 3).33 That high aim is why it is important for a warrior to think carefully before speaking, hard as that is when the group is cheering him on. Once spoken, the warrior’s oath must be carried through; his dom depends on it. The courage and dedication with which he accomplishes the vaunted deed, along with its success or noble failure, builds and determines his identity within the masculine duguđ culture of the hall, the circle of proven warriors. The Wife’s Lament questions this oath culture from the perspective of someone to whom second-level oaths have been made and broken.34 The reader’s part in that questioning concerns the husband’s attitude toward the Wife as manifested in his actions: does he send her to live in the woods because he has become hostile toward her (Klinck, Niles), or is her suffering in that isolated cave-life a by-product of his own necessary, and perhaps temporary, exile (Wentersdorf and others)?35

Before he leaves, the husband broods in “manly” silence, a practice that is not unique to The Wife’s Lament. Here the argument again engages that more famous Exeter Book elegy, The Wanderer. When The Wanderer poet gives voice to his beleaguered protagonist, the very first thing he does is to introduce “the imagery of silence” (Bjork 2008). First the poem’s persona laments that he must speak his woes alone because no one now lives to whom he can open his heart (we find out later that his lord is dead), but next he comforts himself with the truism that it is an indryhten þeaw (“noble custom”, line 12)36 to lock up one’s unhappy feelings, as in a treasure-chest:

Oft ic sceolde āna ūhtna gehwylce

mīne ceare cwīþan. Nis nū cwicra nān

þe ic him mōdsefan mīnne durre

sweotule āsecgan. Ic tō sōþe wāt

þæt biþ in eorle indryhten þēaw

þæt hē his ferðlocan fæste binde,

healde his hordcofan, hycge swā hē wille. (The Wanderer, lines 8–14)

Often I must, alone, every morning

speak my sorrow. There is none now alive

to whom what’s on my mind I dare

openly say. I know as a truism

that it is in a man a noble custom

that he bind fast his spirit-locker,

hold [shut] his hord-coffer, think whatever he will.

With lord and kinsmen all dead in battle, the Wanderer has no social matrix that will allay the pain of his sorrow or hold it at bay. Nevertheless, he is consoled by the socially sanctioned indryhten þeaw of keeping one’s sorrows “bound up” within. More sensible is the preventative measure advised in lines 70–72 (“a man must wait, etc.”, quoted above) that cautions against making an oath until as certain as humanly possible of the ability to honor it, with no regrets. The word beorn places this admonition in a warband context where breaking a public oath would reveal the warrior to be lacking in honor, missing a vital part of what would identify him as a worthy member of that high-status community. The advice to consider “whither the urges of the heart might turn” may hint at the problem of a conflicting commitment that unexpectedly complicates the honorable carrying out of an oath.37

The word collenferđ in line 71 of The Wanderer38 (quoted above) suggests a rashness to be guarded against, even though eagerness for action is a virtue in a warrior. The poet develops an understanding of oath-taking in stages through the poem, first presenting the “stiff upper lip” maxim (about keeping quiet about one’s woes) in lines 11–14, then advising caution in oath-keeping in lines 70–72, and finally adding to these two an important qualifier about knowing the bot before acting, in lines 112–114.39 Bot is an ambivalent word having both secular and religious meanings, like treowe (pledge/religious faith) in line 112a and fastness at the very end of the poem at line 115b (fortress/stability), and the poet takes full advantage of this double meaning of bot. Recent scholarship has downplayed its secular aspect. The gnomic sentence about bot, in emphatic long lines, repeats the word beorn that directs the wisdom to warriors, with lines 112b–113 referring back to the “noble custom” of keeping silent in lines 11–14, and with bote cunne of line 113 echoing cunne gearwe of line 69:

Til biþ se þe his trēowe gehealdeþ; ne sceal nǣfre his torn tō rycene

beorn of his brēostum ācȳþan, nemþe hē ǣr þā bōte cunne

eorl mid elne gefremman. (The Wanderer, lines 112–114a)

Good is he who holds to his troth [pledge]; his grief he must never too quickly,

a warrior (beorn), from his breast make known, unless he first knows the remedy,

the nobleman, how to achieve it with vigor.40

The meaning that most recent commentators assign to the word bot is “remedy”, as translated here, with the idea that the “remedy” for suffering is to put one’s faith in God.41 Within the “heroic” or legal context, however, bot refers to recompense, so interpreting bot “heroically” refers the reader back to lines 70–72, where the beorn is told to be careful about making a commitment until he knows “whither the thought of his breast will turn”. Once committed, to carry out his oath in good faith he needs to be clear about what he must do (the bot) to prove himself loyal to his word and fully a warrior. E.G. Stanley has recently edited The Exeter Book poem Precepts (Stanley 2018, pp. 277–95), in which these lines on the same subject of maintaining silence occur:

Ymb his forđgesceaft, nefne he fæhþe wite,

wærwyrde sceal wisfæst hæle

breostum hycgan, nales breahtme hlud. (Lines 55–57)

Stanley translates: “A truly wise man must reflect on his future in his inmost thoughts, unless he experiences contrariety [line 55b], careful in speaking, not at all clamourously noisy.” Stanley’s note on line 55b betrays the fact that he is not certain about what nefne he fæhþe wite means.42 I suggest that Precepts 55b and the similar phrase at 112b in The Wanderer (quoted above) shed light on each other, and that the “remedy” alluded to in the word bot is an act of violence, fæhþe. Thus I would translate these three lines of Precepts as follows, with bracketed additions for clarity: “Concerning his future, unless he knows the violence [that he must perform],/a firmly wise man shall be wary of words,/ponder [that bot] within his breast, not noisily aloud.” When the husband in The Wife’s Lament is secretively and silently pondering violence under a “blithe demeanor”, he is conforming to a “wise” practice endorsed by his warrior class.43

Thus, lines 112b–113a of The Wanderer, supplemented by the Precepts passage, mark a return to the theme of oath-keeping and the “imagery of silence”, as the bot in line 113b refers to what comes after the oath. However painful or terrifying the task to be done may be, it is best to plan for it in silence until certain about how to carry it out. As Hamlet knew so well, vengeance is a dish best served cold, specifically so that the avenger can take the time to calculate calmly the best strategy for achieving success. This requires the oath-taker to contain within himself the passion that led to his vaunting words, an anger and grief that might concern anything from a minor insult to the murder of a kinsman. The danger addressed by these cautions about oath-taking must have been the easy violence stirred up by manly posturing at a hall celebration as warriors are forced by their “honor” (and mead) into swearing to perform actions that reflection or a change of circumstances might make undesirable. As the “Rad” stanza of the Old English Rune Poem suggests, daylight may reveal the task as surprisingly daunting. In that poem, two different meanings of the rune-name Rad—“advice” (or talk) and “riding” (taking action)—are exploited in a stanza that appears as a maxim: “Rad in the hall for any warrior is easy [sefte, soft], and very strenuous for someone who sits on a powerful horse on the mile-paths”.44 Talking big in a safe place among raucous friends may be all too easy.

The story of the Jomsburg men’s oath-taking vividly illustrates the kind of swearing of oaths at mead that builds bonds while binding the swearer to a future deed (a ritual to which both Beowulf and The Battle of Maldon famously allude, among other works). The Icelandic story of the feast that King Svein held in memory of his father King Harald appears in several versions; the version condensed here is quoted by Einarsson (1968, pp. 109–10) from the Heimskringla account in the Morris and Magnússon (1893) translation of that work. At this feast a great quantity of “grave-ale” is being served—

And that cup of memory must all drink who were at the feast […].

And then were borne to the Jomsburgers the biggest horns of the mightiest drink that was there […].

But thereafter drank Earl Sigvaldi the memory of his father, swearing oath therewith that before three winters were worn away he would come into Norway and slay Earl Hakon or else drive him from the land.

Then swore Thorkel the High, the brother of Sigvaldi, that he would follow his brother to Norway, nor ever flee from battle leaving Sigvaldi fighting.

Then swore Bui the Thick that he would fare to Norway with them, and in no battle flee before Earl Hakon.

Then swore Sigurd his brother that he would fare to Norway, and not flee while the more part of the Jomsburgers fought.

Then swore Vagn Akison that he would fare with them to Norway, and not come back till he had slain Thorkel Leira, and lain a-bed by his daughter Ingibiorg without the leave of her kin.

Many other lords also swore oath on sundry matters. So that day men drank the heirship-feast.

But the morrow’s morn, when men were no more drunken, the Jomsburgers thought they had spoken big words enough; so they met together and took counsel how they should bring this journey about, and the end of it was that they determined to set about it as speedily as may be. So they arrayed their ships and their company; and wide about the land went the fame of this.

In this simple story, these brothers and their loyal friends have no trouble coping with conflicting loyalties. All they must do to “hold to their troth” is to carry out the raping and murdering that they have sworn to accomplish.

In his well-known essay on “The Heroic Oath” (Renoir 1963, pp. 237–66), Alain Renoir shows that carrying out an oath (at least in literature) was not always so unencumbered. Brief summaries of only two of the stories that Renoir tells in that essay will be enough to demonstrate the unanticipated conflicts that an oath-taker may have to face and the amazing rigor with which the oath must be honored despite any personal grief that honoring it may entail. The milder of Renoir’s two stories is about Walter of Aquitaine. This hero’s tale was told in the Old English poem Waldere, but only two fragments of that poem remain, so one must supplement it with scenes from the possibly contemporary Latin epic Waltharius. This is a great adventure tale. Waltharius and his betrothed, Hildegund, both hostages to Attila, have escaped from Hunnish captivity with treasure that Hildegund has cleverly managed to spirit away. Presumably with a horse to carry the treasure and their gear, they take an (estimated) two-month-long hike along the Volga from what is now Hungary to the crossing of the Rhine into the Vosges area of what is now France. However, the ferryman of the Rhine has heard the clink of gold and reports this to the Frankish king Gunthur, who greedily sets off with 12 warriors to take the treasure. The greatest of his warriors is Hagen, a previously escaped hostage of the Huns with whom Waltharius took an oath of friendship when they were captives together. The two groups clash in the Vosges Mountains, and Waltharius takes on Gunthur’s men one by one, killing each in turn until only his friend Hagen is left alive. Hagen has tried all along to talk his obsessed king out of this battle, but to no avail, and when his turn comes to make the final attempt on Waltharius, he must keep his word to the king whom he does not respect against the personal vow of friendship. We cannot know how the Old English Waldere ended, but after Hagen and Waltharius chop each other up until both collapse, this Latin epic concludes with the three survivors, those two and Gunthur (now also maimed), sitting around and chatting about the fight and their wounds. No modern scholar knows what to make of this strangely unheroic ending, though some have very strong opinions.45 In any case, the story makes my point clearly: Hagen’s loyalty oath to his king trumps that to his friend.

Renoir’s second example of an oath to a leader that results in remorse is from the twelfth-century Old French chanson de geste titled Les Quatre Filz Aymon (The Four Sons of Aymon). When Aymon’s oath of service to Charlemagne takes precedence over commitment to family, this story concludes with the tragedy one might expect. Aymon’s sons are oath-bound to the king’s enemy, so under pressure from Charlemagne Aymon must unwillingly fight all four of them to the death. Remorseful though he is, he nevertheless takes pleasure in the valor with which his sons, one after the other, take him on. To me, the situation is unbelievable; why do they not unite in turning on the king, whatever the cost? Again, however, the story makes the point well: these oaths of allegiance to king and comitatus, potentially to be honored even to the point of killing one’s closest friend or dear children, had a power over the oath-taker’s mind and sense of identity that no sane modern person can come close to comprehending.

Few oaths in real life were so fraught as either of these, of course, although the obligation to stand fast by one’s lord at all costs is attested both in the Anglo-Saxon Chronicle and by the poem The Battle of Maldon. In the Parker Manuscript, the chronicle entry for the year 757 records Prince Cyneheard’s surprise attack on King Cynewulf. After his victory, Cyneheard offers to let Cynewulf’s men go free, but they choose to fight to avenge their king to their certain death (Parker Chronicle and Laws: A Facsimile 1941, p. 127; Swanton 2000, pp. 46–47). The Battle of Maldon was composed soon after the event of 991 that it commemorates. In that poem, after the invading Danes have slain the Ealdorman Byrhtnoth and a group of cowardly warriors have fled the battle (lines 185–197), the remaining loyal warriors urge themselves on, one of them reminding the others what we oft æt meodo spræcon, þonne we on bence beot ahofon (“we often spoke at mead when we on the bench raised the oath”, lines 212b–213a). “Now we will find out who is brave,” he concludes at line 215 (Dobbie 1958, p. 13), and they, too, fight on to the death. Carrying out the oath means everything. Renoir attempts to explain the power of the “heroic oath” in respect to this understanding of it by reference to Beowulf:

The nature of the heroic oath in Beowulf, confusing as it may be to the modern scholarly reader, posed no serious problem for the protagonists of the poem. Even the jealous Unferth, who insultingly questions Beowulf’s adequacy to deal with the monster Grendel (lines 499–528), never openly doubts the intrinsic value of the latter’s oath to rid the world of the evil doer. The only problem is the fulfillment; if Beowulf either kills Grendel, and later his mother, or bravely dies in the attempt, he will have fulfilled his oath to the letter and won eternal renown among both the Danes (lines 660b–61) and the Geats (lines 1484–86).

(Renoir 1963, p. 242)

It is argued here that when the husband in The Wife’s Lament breaks the vow never to part that he and the speaker took together (wit beotedon, line 21), he must have been under a compulsion to carry out a “heroic” oath similar to that which caused the real-life companions of Cynewulf and Byrhtnoth to fight on after their leaders’ death, that caused Beowulf to swear to do or die and even heroically to add odds in the huge monsters’ favor (he would bear no sword against Grendel; he would go alone against Grendel’s mother and later the dragon); and that caused the legendary Hagen and Aymon, under oath to their respective kings, to take up arms against those they cared for. Despite some recent readings of The Wife’s Lament that claim otherwise, the Wife says nothing about her husband turning against her; she speaks only of the desire of his family to part them, so the easiest scenario to imagine from the inconclusive hints we are given—the man’s dark brooding about violence under a “blithe demeanor” and his going “over the waves” to who-knows-where—is that he was compelled to take action abroad by an oath more substantial in the culture of the poem than a vow made to a beloved woman.46

However, in evoking an imagined future, oaths share a semi-magical proleptic force with cursing and with this element comes a danger in the breaking of them that has not been suggested before in discussions of Old English poetry (so far as I am aware). The result of not carrying out to the letter an oath made with Big Words may be even worse than loss of honor. In the mind of a warrior, always conscious of what might hinder or help at need, a serious oath that was “sanctified” by raising the meadcup and then left unfulfilled might be thought to turn the psychic force of that oath back upon its maker, perhaps awakening a malevolent reprisal as powerful as a curse. Giorgio Agamben refers to such forces (Agamben 2011, p. 12) when he quotes the historian of Greek law Louis Gernet, who speaks of the ‘sacred substance’ with which the one who utters the oath is put in contact: “To swear, therefore, is to enter the realm of religious forces of the most fearsome sort” (cp. Gernet 1981, p. 223).47 If the husband in The Wife’s Lament even subconsciously feared reprisal for oathbreaking by “forces of the most fearsome sort”, that superstition might have led him even more profoundly than the need for dom in his choice to break a less binding vow to his wife.

However, he did have a choice about what to do about the wife whom he must leave. Assigning him a caring concern about her safety is speculative (though less so than assigning him hostile intent), but if the poet locates the Wife’s story in a pre-Christian period, then what he does makes sense. In that pagan world, the husband would know that the sanctuary of a herheard was an excellent place to keep a woman safe. According to recent discoveries about inhabited caves in the discussion that now follows, the poet may well have been familiar with some acceptable havens in his actual ninth- or tenth-century landscape, locations to which the husband of his poem might feasibly entrust his wife, probably with a servant who would come to look after her daily needs.48 As Niles suggests in God’s Exiles (Niles 2019, p. 237), by labeling her haven with the word herheard that identifies it as a pagan temple site, the poet locates the lonely Wife’s underlying narrative in an imagined pre-Christian world of long ago.

3. The Caverns under an Oak Tree as a Retreat

Few would disagree with Anne Klinck that, whatever framework we choose for interpreting it, The Wife’s Lament

remains something of an enigma. The speaker’s feelings, not the events of her life, nor even in any exact sense her physical surroundings, are the focus, and this makes the poem both highly evocative and at the same time tantalizingly laconic and elliptical. Probably the poet intended to mystify; the woman’s circumstances are disturbing largely because they are strange and undefined.

(Klinck 1992, p. 54)

Despite this lack of definition, some new ideas are offering a fresh perspective on the context of The Wife’s Lament, suggestively in recent revisionist work on The Exeter Book by John D. Niles (2019) and his student Brian O’Camb (2009), and more denotatively in assessments of recent archaeological excavations of cliff-side dwellings and human-altered caves—one of which may plausibly have sheltered a named ninth-century king (discussed below). While the Wife may find her current habitation strange and perhaps threatening to her status, thereby leaving her undefined and in mental turmoil, if those “caves” in which she has been ordered to live are anything like the “rock-cut buildings” in the English Midlands that are attracting current attention, her dwelling place may not be so unpleasant as some have assumed.

Michael Lapidge claims that “the poet is describing a mental landscape, not a physical one; in its loneliness and desolation it is a visible embodiment of the narrator’s invisible grief”. (Lapidge 1997, p. 34; Emily Jensen goes so far as to say that there is no actual cave at all; Jensen 1990, pp. 450–51). Though we will challenge his assertion that the poet does not intend to evoke a physical landscape, Lapidge is correct in saying that the speaker’s representation of her surroundings is colored by her grief. It should be observed, however, that the Wife only complains about the looming landscape around her and the isolation and loneliness that she feels; she does not complain about the physical nature of her cave-dwelling, never suggesting, for instance, that it is damp and cold. It is her isolation in a land far from her home and friends (see line 16) that makes that place a wic wynna leas (“joyless dwelling”, line 32). Because it is called an eorđscræf (line 28, usually translated “cave”), modern readers of the poem have assumed that the poor woman has been forced to live in a dank and horrible hole underground, as in this passage from Burton Raffel’s popular interpretive translation that vaguely reflects lines 28 and 34–35 of The Wife’s Lament:

I writhe with longing in this ancient hole […]

While at dawn alone, I crawl miserably down

Under the oak growing out of my cave.

(Raffel 1998, pp. 14–15)

In fact, there is reason to believe that (in the poet’s imagination) her husband has arranged for the Wife to occupy a clean and well-lit sanctuary cave structured like one of the many medieval and later hermitage caves under the care of a monastery, where her needs would be looked after much as if she had chosen isolation for religious purposes. The meager hints offered in the poem allow us to suggest that the Christian poet knew, possibly from personal experience, although hearsay would be sufficient, of livable “hermit” caves associated with monasteries, while locating the cave of the poem in a long-ago time that Niles describes as “a raw and primitive past” (Niles 2006, p. 206).49 By having the Wife call the site a herheard (line 15) without displaying fear of its heathen aspect, and calling the cavern “old” (line 29), the poet invites us to imagine her in an ancient native place of sanctuary that in later times might be appropriated and administered by a Christian community.50 Even the oak, twice mentioned, that stands on the hill above the cavern suggests a pagan location—although oak trees often stand tall on hilltops quite naturally.

The Wife’s cavern, pre-existing in the landscape and ready for habitation, is not unique within the context of early English poetry. It is similar to the cave found by Lot in his escape with his family in the Old English Genesis. That dwelling, too, is called an eorđscræf at line 2597:

Ne dorste þa dædrof hæle

for frean egesan on þam fæstenne

Ieng eardigean, ac him Loth gewat

of byrig gangan and his bearn somed

wælstowe fyrr wic sceawian,

ođþæt hie be hliđe heare dune

eorđscræf fundon. Þær se eadega Loth

wærfæst wunode, waldende leof,

dægrimes worn and his dohtor twa. (Genesis, lines 2591–99; Krapp 1931, p. 77)

The valiant hero dared not then/for fear of the lord, in that stronghold/dwell longer, but he, Lot, departed/to walk away from that city [Sodom], and his children with him,/far from that death-place, to seek a dwelling,/until by the slope of a high hill/they found an earth-cave. There the fortunate Lot/dwelt faithfully, the beloved ruler/for many a day with his two daughters (translated by the author).

Drawing on Paul Battles’ argument in “Of Graves, Caves, and Subterranean Dwellings: eorđscræf and eorđsele in The Wife’s Lament” (Battles 1994; cp. Harris 1977), Lori Ann Garner is persuaded that the word eorđscræf at Genesis 2597 refers to a souterrain, and it is true that the words eorđscræf and souterrain both refer to a dwelling “scraped” (scræf) from the earth, thus not a natural cave. Whereas the word eorđscræf is directionless, the etymology of souterrain emphasizes a downward direction. The possibility of a uniquely English horizontally cut cliff dwelling occurs to neither Battles nor Garner, as no scholar was yet aware of the cluster of such buildings in the Midlands (discussed below).51 Of all suggestions made, however, their idea of a souterrain comes closest to the building type proposed for the Wife’s cave in this essay. In connection with Lot living in an eorđscræf with his dohtor twa (“two daughters”, line 2599), translating the Latin filiae eius (19:30), Garner adds a minor point worth considering in the Wife’s context: “Souterrains often function as a protection and hiding place in Germanic literature for females more widely” (Garner 2011, p. 173). Her comment recalls another female cave-dweller speaking her plight in The Exeter Book only a small block of riddles away from The Wife’s Lament. The speaker of Exeter Book Riddle 15, identified as a vixen by Audrey L. Meaney (2000), plans actively to alter her earthen dwelling when, threatened by an enemy breaching her “souterrain” (my term here), she will dig a new path through the hillside (lines 17–18) to create a “secret way” out (line 21).52

But Garner finds that the word burgtunas does not fit well with that souterrain concept in The Wife’s Lament:

Potentially problematic to conceptualizations of such a space as either a barrow or a souterrain is the phrase bitre burgtunas in line 31. The compound burgtunas, which literally means something like ‘fortification towns’, is unique in Old English poetry, and it is difficult to reconcile the woman’s apparent underground solitude with the presence of such a grand space, with the designation bitre (“bitter”) only adding to the quandary. Leslie offers “protecting hedge” as a gloss and suggests it is “possibly an ancient earthwork” (56). Leslie further argues that bitre is used here in the sense of ‘sharp’; thus describing “briars which have grown over the protecting walls of the cave or mound, although the abstract meaning ‘bitter’ may be intended as well” (56). Even more specifically, Klinck posits that “the bitter enclosures overgrown with briars” may be the remains of an abandoned settlement.

(Garner 2011, p. 184)

Shari Horner argues that these burgtunas are not abandoned because the Wife’s situation should be associated with that of an enclosed nun near a monastery (Horner 2001, p. 52). Even if one objects to making the Wife into a sort of proto-nun, Horner’s idea of the woman being enclosed in a site under the guardianship of a monastery may offer a clue to the type of dwelling in which she finds herself—in the poet’s imagination. Two other poems in the Exeter Book, Guthlac A and Guthlac B, attest to knowledge of the story about the hermit of that name whose place of “isolation” in a known location had become, by the time of The Exeter Book poems, relatively near a monastery. Soon after 700, the young nobleman named Guthlac (a real person) chose to become a hermit in the wilderness near Croyland in what is now southern Derbyshire. His vita by a near-contemporary named Felix of Crowland, writing around the mid-eighth century, tells how Guthlac dug into the side of a barrow “built of clods of earth” and built his hut in that curious location. It was isolated but contained a human-built structure (the barrow) and could be visited, although nearby Croyland Abbey that ministered to that hermitage was established later.53

The Wife’s wilderness home seems different from Guthlac’s in several respects, wilder and more genuinely isolated, with a shelter more like that of the eorđscræf taken over by the dragon in Beowulf. But that is only an impression because the poet does not provide enough information to allow the reader to envision her cave in the detail that the Beowulf poet provides for the dragon’s lair. In exploring that description in Beowulf, Niles refers to a series of passages that include six eorđ-headed compounds, four different words with one of them repeated: earth-building, earth-hall (three times), earth-scraped structure, and earth-wall. As Niles explains:

What Beowulf sees there is the opening to an earthen hall or chamber (eorđreced 2719a; cf. eorđsele anne ‘a singular earthen hall’ 2410; cf. also eorđse(le) 2232a, eorđsele 2515a, eorđscræf 3046a, eorđweall 3090a). The chamber is roofed by stones supported by upright pillars (the vaulting is stapulum fæste ‘firm on its pillars’ 2718b). The ceiling (inwithrof ‘inimical roof’ 3123b) is high enough that a person may walk about inside unimpeded (2752–2755, 3123–3125; cf. 2244–2246a, 2268b–2270a), while the room is capacious enough to contain a large treasure hoard plus a fifty-foot-long dragon (coiled up, one assumes). Eight men can occupy the chamber concurrently (3120–3124a)

(Niles 2012, p. 27).54

Despite the more meager description of the Wife’s herheard, there exist promising analogues for her cave. Amid the cluster of rock-cut dwellings in the Derbyshire Midlands (in roughly the same area of England as Guthlac’s dwelling) are cave complexes with interiors large enough to allow someone like the Wife to pace about in them, as she does in her frustration and growing anger, walking “alone at dawn through these earthen caverns under the oak tree” (lines 35–36).

4. Dido’s Fury and the Wife in Her Cave

This anger takes us back to Queen Dido, who believed that her night spent with Aeneas in a sheltering cave meant that they were betrothed and that in leaving her forever he betrayed that troth (Aeneid 4: 316–30, 431).55 Dido’s perception of betrayal leads to the powerful curse that she launches at the departing Aeneas before killing herself with the sword that he gave her as a love gift. There is no hint in The Wife’s Lament that the Wife is as desperate as Dido, and in fact the final word of the poem, the verb abideđ (“awaits”), hints at an expectation that her husband, if he survives, will come back to her. Nevertheless, her anger at being left alone in the cave crescendos into a curse that he may experience a similarly aching isolation. Thus, it is the cave experience in each poem that arouses the pagan fury of both Dido and the Wife, inciting each high-status woman, humiliated by being abandoned, to curse her felaleof, her much-beloved man.

However, the cave itself is not the object of the Wife’s fury, and indeed she probably could not justly complain about its discomfort. When elite exiles and even royalty retired for a time to dwell in a hermitage “cave”, they probably had comfortable furnishings and dutiful minions to serve them—completely unlike the “place of hellish exile” imagined for the Wife by both Alaric Hall (2007, pp. 224–32) and Sarah Semple (2013, pp. 152–53). In what follows, this essay will introduce examples of real “caves”, some even situated “under an oak tree”, that may represent the sort of building the poet may have known about and envisioned as a safe house for the lonely Wife. Precise dating of such buildings is impossible because they may or may not have been dug out from what was initially a natural cave, and the soft stone that makes their construction attractive also makes it easy to alter, so that the date of the dwelling becomes fluid. Nevertheless, certain features within these buildings suggest that they may be dated at least to an era because of stylistic correspondence to similar features in more ordinary buildings. That such buildings were known in the days of King Alfred (848/49–899) is indisputable because of an aside by Asser in his life of the king written in 893 (Keynes and Lapidge 1983, p. 53). In telling how the Great Army of the Vikings left Northumbria and came to Nottingham in 868, Asser (a Welshman) breaks off to remark that the city” is called Tig Guogobauc in Welsh or Speluncarum Domus in Latin” (Keynes and Lapidge 1983, p. 77). Asser’s editors remark that the Welsh place name “does mean precisely ‘cavy house’ in Old Welsh”, but they then suggest that the name “may perhaps have been Asser’s invention, based on his own knowledge of the place” (Keynes and Lapidge 1983, p. 241). The City of Nottingham, built on soft sandstone in the same general Midlands area as Repton Abbey and the rock-house cluster, is riddled with rock-cut buildings (“approximately 840 known medieval and post-medieval man made caves”56), mainly used for storage and brewing in recent times, and now mostly abandoned. It is only Asser’s odd naming of the place that allows us any certainty (beyond their physical existence) that rock-cut buildings in that Midlands area were noticed at all before the modern era. An important aspect of these cave-like buildings is that, unless they are spectacular in some way, they tend to go unnoticed, and once untenanted may fall to ruin, may revert to existing as “mere rock”.

The “cave under an oak tree” is no ruin. Asser’s fleeting aside with its implicit ninth-century date allows us the pleasure of imagining the desolate young Wife weeping her summerlong day outside of an earđscræf that might look something like Blackstone Cave in Worcestershire, as shown in this 1721 copper engraving by Elisha Kirkall for William Stukeley’s Itinerarium Curiosum series (see Figure 1). The location and number of windows carved in the cliff-face suggest a generous interior with more than a single chamber. Today a surrounding grove of trees and other hindrances make access difficult.57 The following section is the British archaeologist Edmund R. Simons’ account of current excavations of such rock-cut buildings as this.

5. Current Archaeology of Rock-Cut Buildings

A Contribution by Edmund R. Simons, MSt, IHBC, FRGS, FSA

The current archaeological project titled “Rock-cut Buildings of the English Midlands” is a preliminary attempt to identify and assemble a corpus of rock-cut buildings that will eventually extend throughout the UK.59 (Although the present project focuses on the Midlands cluster of these buildings, important outliers exist in almost every corner of the country.) These rock-cut buildings are not natural caves; almost all those considered in this project are entirely artificial structures, generally cut into soft sandstones, limestones, or tufa. A few such sites were occupied as dwellings as recently as the 1960s, and many are still in use as sheds, cellars, and outbuildings. Despite some very limited antiquarian interest during the eighteenth and nineteenth centuries,60 and despite the vast date range and the huge number of these structures, with over 400 in the Midlands alone, they have previously been studied only in isolation. This ongoing project attempts to begin to rectify this situation by regarding these buildings as a type and creating a draft typology to develop, test, and apply techniques to analyze the negative stratigraphy, to understand the archaeology of voids, and to develop techniques for analyzing and phasing of rock-cut buildings. It is the first attempt to understand how, when, and why these structures were built, how they were used, and how they change over time—that is, to understand them in the same way we would understand any other building type.

The early results of studying these buildings as a type rather than individually have been something of a surprise, as roughly half of the dwellings looked at in the principal study area (the Permian and Triassic sandstone counties of Worcestershire, Shropshire, Staffordshire, Derbyshire, and Nottinghamshire) can be credibly suggested to have a medieval or earlier origin.61 A good case can be made that some of the large multi-room rock-cut sites studied in the project were in use or partly constructed before c1200. These include (among others) the following major sites:

• Anchor Church (Derbyshire),
• Dale Abbey Hermitage (Derbyshire),
• Lenton Hermitage (Nottinghamshire),
• Bridgnorth Hermitage (Shropshire),
• Redstone Hermitage (Worcestershire),
• Blackstone Hermitage (Worcestershire),
• Southstone Hermitage (Worcestershire),
• Crachcliffe Hermitage (Derbyshire),
• Guy’s Cliffe (Warwickshire).

Immediately apparent from this list is that all of these sites are known as or inferred to be religious hermitages. All have known or credibly demonstrate later medieval (twelfth to sixteenth century) activity that survives in the historic and archaeological records, or they contain diagnostic medieval features such as pointed arched doors, smoke hood fireplaces, quatrefoil windows, etc.62 Moreover, several of the entirely rock-cut sites have very strong evidence of prehistoric activity, with periodic use over very long periods of time.63 Of particular interest when thinking of the “cave under an oak tree” in The Wife’s Lament are the many structures with substantial standing remains that have documented histories and diagnostic features similar to those found in conventional early medieval English buildings. In what follows, I will first trace roughly what is known of the medieval history of British “cave” dwellings (broadly considered), beginning with the Celtic practice and then moving to the current study of English sites. This will lead to consideration of the probably early medieval English period sites and the possibility of their providing a context for the poet’s imagining of a “cave under an oak tree” in The Wife’s Lament.

The use of caves in Britain as places of sanctuary, exile, reflection, and solitude, and simply as habitation, goes far back in antiquity. As elsewhere in the world, many British caves have been a focus of spirituality and ritual for millennia; they include Paleolithic burials, Bronze Age ritual use, and Iron Age and Roman shrines. The sites are numerous, and many of the earlier ones are well-understood.64 What is less understood, and has been comparatively neglected in archaeological study despite considerable historical evidence, is how this ancient ritual use of cave sites crossed over from a broadly pagan context into revised use in the Christian era.

There does exist, however, contemporary and later historical evidence for use of caves in post-Roman Britain and Ireland. This documentary evidence almost always relates to religious figures, many of whom were later to become saints or were recognized as such in their lifetimes. In “Saintly Associations with Caves in Ireland from the Early Medieval Period (AD 400–1169) through to Recent Times”, Marion Dowd describes how, during the early medieval and medieval periods, the prehistoric human usage of caves as funerary sites shifted to encompass a wider variety of secular and religious uses. Piecing together local folklore, place names, and traditional Christian religious practices such as pilgrimages celebrating local saints, Dowd hypothesizes how early medieval ascetics likely used caves (Dowd 2018, p. 125). The discomfort of these caves likely added to their appeal, as Irish monasticism “consciously followed the ascetic and meditative ideals of the desert fathers” (Fisher 2001, pp. 2–3). Early Christians highly regarded caves as suitable settings for solitary contemplation (Dowd 2018, p. 127), and numerous British saints are thought to have lived in caves.65 For example, in his Ecclesiastical History, Bede describes Saint Ninian, a British bishop who was responsible for bringing the Christian religion to the southern Picts (Colgrave and Mynors 1969, p. 223), and the anonymous eighth-century Miracula Nynie Episcopi tells how Saint Ninian lived “in a cave of horrible blackness” (MacQueen and MacQueen 2005, p. 101; cp. Levison 1950).

The cave was traditionally, therefore, part of the suite of places where a hermit could seek salvation, either as a solitary or as part of a community, and for some the very fact of a cave’s previous association with paganism was an attraction. Several saint’s lives include descriptions of the saint’s use of a cave, perhaps to escape supplicants, or as a Lenten hermitage. There are also many locations where the place name is evidence for an early cave-anchorite site, or where historical and folklore associations suggest a connection with a saint, although very few of these sites have been studied or excavated, so we know little or nothing of how, or even if, the saint that legend associates with such a site actually used it. Examples of known or suspected early cave dwellers in the British Isles are usually Celtic, with the majority of the earliest known or suspected saintly cave dwellings found in Wales, Cornwall, Scotland, and Ireland. Persons dwelling in these sites include the important fourth- and fifth-century saints Ninian, Illtud, Samson, Kevin, and Patrick, as well as less well-remembered saints such as Melangell, Serf, and Gwyddafach. Caves long associated with these names are usually natural, small, and simple in the extreme, but they may have some rock-cut features or conventional walling.66 Their nature suggests the search for solitude in wild places.67 This geographic spread of “mystical” Celtic cave-dwelling must be balanced against the physical fact that many of the caves in Britain and Ireland lie in the north and west, where the geology and topography make cave dwelling possible.

It is clear that even as early as the fifth century the use of caves by both male and female Christian solitaries was a well-established practice, and Gildas himself refers to “caves and consolations of holy men” (Gildas 1978, p. 102). Already, however, the actual use of the cave is varied. A familiar pattern of use features a single divine originally seeking solitary prayer but accruing fame that attracts followers. These newcomers may dwell in the cave with the saint for a time, but lack of space (presumably) leads to the creation of a monastery nearby. Saint Melangell, a seventh-century female Welsh saint of supposedly noble Irish descent, is a good example of this pattern. She is associated from the eighteenth century with a rock shelter that bears her name and may be where she lived in solitude before the founding of the monastery.68 Another variant is when an already established holy person with followers seeks periods of solitude or goes to live in a smaller and simpler community. Some such persons may have high-status religious standing or even be bishops, like Illtud, Samson, and Dubricius, all of whom are reputed to have taken periodic sabbaticals in caves.69 As suggested above, for some, the lure of caves is their inhospitality and the difficulty of living within them. Cave dwellings like Saint Kevin’s Bed in Ireland are little more than person-sized shelves in the rock (in this case artificial), capable of keeping off the worst of the weather. In this case, mortification was the key to the choice of the site, but it must not be presumed that this was always the case. The simple convenience that a cave may have presented to a person wishing to live alone, or within a small community having a readymade building, may have often played just as much a role as the quest for mortification. There is a long-running concept, in some cases accurate, that earlier saints become identified as hermits in the twelfth century as a way of retroactively enhancing that saint’s reputation. In recent years, the work of Marian Dowd (2018) and John Blair (2006), among others, has substantially changed this picture to suggest that paucity of preconquest written evidence has hidden the true number of Celtic and early medieval English hermits.70 The cave sites associated with the earliest hermits often follow a similar pattern. They are almost all entirely natural and, although some have been enlarged or have been furbished with stone walls or roofs, or even have rock-cut furniture, their capacity for enlargement is severely limited by the geology of the enclosing stone.

The archaeological project has identified a number of probable early medieval hermitages that are cut into the softer, more easily worked stones, such as the Permian and Triassic sandstones found throughout much of the English Midlands, as well as in parts of Cumbria, Southern Scotland, Northumbria, Sussex, and Dorset. These sandstones are generally poor in natural caves,71 so these wholly manmade rock-cut structures have not been a focus for cave archaeology or studied as a group before now.

Gaps in the chronology remain in these core territories of England, but (as mentioned above) notable early hermits such as Guthlac, Cuthbert, and Chad are all associated with subterranean living in semi-artificial sites, and there may well have been a continuation of the use, creation, and enlargement of rock-cut sites throughout the early medieval period. By the ninth century they were common enough that, as noted above, the Welsh Bishop Asser could describe Nottingham, with its many caves cut from sandstone and thus artificial rather than natural, as “Tig Guocobauc” (“City of Caves” in Welsh). A recent paper by the author suggests that Anchor Church (see Figure 2 and Figure 3), a complex rock-cut site in Derbyshire, may be an almost intact hermitage of the same period and perhaps served as the temporary dwelling place of Saint Hardulph, a deposed king known to historians as Eardwulf of Northumbria (fl. 790–c. 830) (Simons 2021, “Anchor”). Another hermit, a brother of King Athelstan, is said to have retired to the already existing rock-cut hermitage at Bridgnorth (Shropshire) in the early tenth century, an association that may be supported by the rock-cut remains of an early Romanesque chapel with chancel arch of probable Saxon origin.72 Guy’s Cliffe in Warwickshire, associated with the legendary Guy of Warwick, is recorded as already existing by the 920s and was substantially enlarged over succeeding centuries (Clay 1914, p. 31). By the time The Wife’s Lament was written down in the tenth century, the concept of the use of caves for retreat and religious exile was very well-established throughout Britain. The practice was not only familiar from (and thus endorsed by) biblical texts and the lives of desert and continental saints, it had been known in Britain for several hundred years and was undergoing something of a revival.

Those of us involved in the “Rock-cut Buildings” project are currently in the initial stages of understanding the physical remains of cave hermitages themselves, as the earliest features in a complex rock-cut structure can be difficult to define. By making use of a modified version of the archaeological analysis of traditional buildings, however, we are able to identify and define primary spaces and features. This is partly made possible by the preservational qualities of artificial caves, which, unlike conventional buildings, can survive for many centuries unless the cliff is entirely quarried away or completely collapses. Rock-cut buildings can be added to and altered, of course, but deep excavation is a dangerous and intimidating process, so an awareness of such danger by potential “improvers” may have aided the survival of primary features.

The construction dates of the Midland sites chosen for analysis range from the (possible) prehistoric to the 1840s. The earliest legible structures in the sequence are tiny cell-like rooms, sometimes incorporating an enlarged natural hollow or a cleft in the rock. Several sites, such as Guy’s Cliffe, Bridgnorth, Redstone Rock (see Figure 4), and Holy Austin Rock, include raised entrances accessed by ladder or steps; very small, narrow arched-headed doors and very low ceilings are typical features. At other sites we see similarly raised entrances, but with suites of tall irregular rooms above, and these often have slots for timber external walls on the cliff face, walls that later are usually replaced by stone.73 These sites can be large, with identifiable chapels, cells, etc. Some remained in use until the Reformation, and they can be rich in later medieval features such as round chimneys, smoke hoods, large fireplaces, and pointed arches, but analysis of the sequence shows that these features were often inserted into older spaces. Early diagnostic features are few, but they include those details mentioned above at Anchor Church, with tall and narrow arches, tiny elevated arched windows, and simple pilaster heads of a type that appears to predate 1200. Several sites show distinctive triangular-headed niches, which, if found in a conventional building, would be dated as early medieval English with no debate.74 Simple incised crosses and other graffito of early medieval form appear at a number of sites. There is thus strong evidence to indicate that some of these very large hermitages may be of mid- to late Saxon date and may even have earlier origins.75

One of the largest and most complex hermitages, long a place of local pilgrimage and under investigation by the current project, lies at Guy’s Cliffe just outside of Warwick. Possibly dating to the tenth century (with an association since the twelfth century with the legendary Guy of Warwick, hence its name), it is clear that the complex of buildings has grown up around an early cell that is partly excavated into a natural hollow in the rock. This primary cave is surrounded by numerous other rock-cut rooms and tunnels and topped with a fifteenth-century chapel. There has been much fifteenth- to eighteenth- and nineteenth-century work on the site, but the primary cell itself remains largely untouched, as do a great number of possibly early rock-cut rooms. This Guy’s Cliffe site will soon be the subject of scientific investigation and excavation,76 giving us an opportunity to examine, in far greater detail than has been done before, a rock-cut site ostensibly contemporary with The Exeter Book. The size of some of the case studies makes it possible that they had facilities for receiving guests and supplicants. Later records of hermitages with several buildings, including cells, oratory, workshops, and buildings or rooms to receive guests, suggest a permanent staff of servants and other monks or nuns.77

In summary, the initial findings of the current project provide ample evidence for the existence of preconquest rock-cut hermitages that would have been familiar to Exeter’s cosmopolitan and well-connected monks because of their value as destinations for quiet contemplation and potential royal patronage.78 Several details in the The Wife’s Lament chime with locations in the Midlands; for example, the concept of an “earth-hall” may imply that the material is more earthen than stone and is legible as a hall rather than a small irregular cave. The Wife also moves freely through her dwelling, which does not tally with the tiny single-entrance spaces of a reused barrow or chambered tomb like the one described in Felix’s Life of Saint Guthlac (Colgrave [1956] 2007, pp. 93–95). In other words, the poem’s description fits several of the probable near-contemporary rock-cut sites better than it does a natural cave or barrow. Perhaps tellingly, at least three sites on Kinver Edge, Holy Austin Rock, Nanny’s Rock (see Figure 5), and Solcum Aylesbury, also have oak trees growing above caves that are positioned immediately under the defenses of hillforts reused in the sub-Roman period, and oaks similarly tower above Anchor Church, which sits near Viking fortifications.

It is plausible that the Wife’s location is a former pagan site, whether we imagine her in a natural cave or a rock-cut structure. Some rock-cut buildings investigated in the project show pre-Christian usage or construction. If we assume her to be a lady of considerable social standing, abandonment in a natural cave seems improbable. In Coastal Landscapes and Early Christianity in Anglo-Saxon Northumbria, David Petts notes that since many hermitages were visible from areas connected with the monarchy, they were included in a network of royal protection and patronage, even if they were geographically isolated. High-status male and female guests were not only permitted to see and visit such locations but even inhabited them on occasion (Petts 2009, p. 88). John D. Niles posits that The Exeter Book poems serve in part as “a means of acculturating high-ranking members of Anglo-Saxon society to cenobitic ways of life” (Niles 2019, p. 13). In that case, it is likely that a tenth-century monastic poet would have been familiar with the concept of established cave hermitages as ideal locations for prayer and solitude. It is possible that the poet’s inclusion of such a site was an effort to increase interest in eremitic living among high-status individuals at a time and place where hermitages functioned as a component of a broader cenobitic entity.

6. Conclusions (by the Author)

“Rock-cut”, i.e., human-made, cave dwellings are not mentioned in books on early English archaeology and are absent from their indices for one clear reason: no indisputable archaeological proof exists that any extant rock-cut building known was created that long ago, no coin or brooch has been recovered, as yet, to give evidence of a likely date. The only documented record of an awareness of such buildings is Asser’s Welsh name for Nottingham, and the only archaeological evidence consists of some suggestive stylistic features at certain Midlands sites. Yet, it seems entirely plausible that the poet of The Wife’s Lament adopted a cliffside hermitage, known in person or by hearsay, to evoke the image of a complex of rock-cut chambers where a young woman lived, lonely and perhaps furious, but safe.

The reuse of some of these buildings in modern times makes the point clearly that the Wife has not been sent to suffer in a hole beneath the roots of an oak tree. However miserable and furious she may be, if she has been sent to a rock-cut building like some of those described above, she is well-domiciled. The poet who placed her there in his imagination was clearly not concerned that the eorđscræf might be inappropriate lodging for a young lady of high rank, but he might nevertheless have been surprised at how such “caves” could be recovered for reuse a millennium or more later. For example, rock houses carved into Kinver Edge, a sandstone cliff near Stourbridge in the West Midlands, were inhabited until the mid-twentieth century, and some are now being restored to that period by the National Trust. According to the Trust, “The most famous are the homes at Holy Austin Rock, now restored and open to visitors” (Kinver Edge and the Rock Houses n.d.). These buildings are cozy and quite pleasant, with a functioning tearoom on the upper level. Another site, the “Rockhouse Retreat” in Worcestershire, provides a more startling example. The owner, claiming the building is medieval, rents its refurbished rooms to tourists for GBP 231–286 (USD 333–379) per night (The Rockhouse Retreat 2021). These two renovations demonstrate that rock-cut dwellings, when whitewashed and well furnished, could provide surprisingly luxurious accommodation, suitable as a safe house for a high-status wife in (imagined) pagan times, as a comfortable retreat for an exiled ninth-century king, or as an unusual holiday getaway for a wealthy modern tourist.

This essay regards as dated the idea that the husband in The Wife’s Lament was “cruel” in sending the Wife of the poem to live in an eorđscræf under an actreo. Current archaeology suggests that she could have been comfortable and well looked after in such a site, most likely a spacious rock-cut building, perhaps like one in use at the time the poem was written that the ninth- or tenth-century poet relocated in imagination to a “pagan” time of long before, calling it a herheard.