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

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.


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 Lament 1 is a poem in the voice of a young woman 2 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 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 geheaderod, "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 faehdu 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 faehdu ("the feud[?] of my much-beloved"). Klinck translates faehdu "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 faehdu 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 faehdu as "hostility", the more general meaning now assigned to that word, mines felaleofan faehdu, 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 (faehdu 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 her 23 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 gebaero ("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 blide gebaero 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 blide gebaero 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 blide gebaero 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. 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 context 29 ). 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.

Speaking Big Words: Toward an Exploration of Problematic Oath-Taking
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 Humanities 2022, 11, 69 7 of 32 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 aeghwylc sceal ende gebidan worolde lifes; wyrce se þe mote domes aer deaþe; þaet bid drihtgum [an] unlifgendum aefter 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 lofdaedas, and lof is the storied record of deeds performed well: "We have heard", says the Beowulf poet, huda 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 dugud 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 sceoldeānaūhtna gehwylce mīne ceare cwīþan. Nis nū cwicra nān þe ic him mōdsefan mīnne durre sweotuleāsecgan. Ic tō sōþe wāt 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 collenferd in line 71 of The Wanderer 38 (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 brēostumācȳþan, nemþe hēǣr 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 fordgesceaft, nefne he faehþe wite, 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 faehþe 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, faehþe. 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 oathbound 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 aet meodo spraecon, 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.

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 eordscraef (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 eordscraef at line 2597:  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: eordscraef and eordsele in The Wife's Lament" (Battles 1994;cp. Harris 1977), Lori Ann Garner is persuaded that the word eordscraef at Genesis 2597 refers to a souterrain, and it is true that the words eordscraef and souterrain both refer to a dwelling "scraped" (scraef ) from the earth, thus not a natural cave. Whereas the word eordscraef 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 eordscraef 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 eordscraef 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 eord-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 (eordreced 2719a; cf. eordsele anne 'a singular earthen hall' 2410; cf. also eordse(le) 2232a, eordsele 2515a, eordscraef 3046a, eordweall 3090a). The chamber is roofed by stones supported by upright pillars (the vaulting is stapulum faeste '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).

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 abided ("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 ninthcentury date allows us the pleasure of imagining the desolate young Wife weeping her summerlong day outside of an eardscraef 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.

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,

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: 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 Figures 2 and 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 midto late Saxon date and may even have earlier origins. 75 Humanities 2022, 11, x FOR PEER REVIEW 23 of 33   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(Colgrave [ ] 2007. 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. Figure 4: Redstone Rock Hermitage. Redstone Rock Hermitage, Worcestershire, is a massive complex of caves comprising at least 10 main rooms and with a known later medieval history and features that suggest an early medieval origin. Photograph E.R Simons.

Figure 5:
Nanny's Rock, Kinver, Worcestershire, is a rock-cut site with early medieval features similar to a number of known or suspected hermitages. The site is made up of a suite of three large rooms along a cliff face and formerly accessed by a stair, with a small, detached cave to the north. Photograph E.R Simons. 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.

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 eordscraef 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 eordscraef 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.

Funding:
The research on Section 1, Section 2, Section 3 and Sections 4 and 6 received no external funding. Section 5 (A Contribution by Edmund R. Simons, MSt, IHBC, FRGS, FSA) was generously sponsored by Professor Richard Skinner of Houston, TX, USA.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.

Data Availability Statement: Not applicable.
Acknowledgments: My thanks go first of all to two long-term friends, Robert E. Bjork, who engaged me in this project, and John D. Niles, who invited me to join him in a previous project combining early English literature with current archaeology, resulting in Beowulf and Lejre (Niles and Osborn 2007); Niles' recent work inspired this essay's range and format. Scott Gwara and Leonard Neidorf generously sent me copies of their articles and books, abundant with fresh ideas that enriched the oaths essay. The anonymous peer reviewers for the press contributed further ideas and improvements. The archaeologist Edmund Simons leaped in with enthusiasm when invited, adding an interdisciplinary real-world perspective that raises discussion of the Wife's "cave" to a higher level. This article might never have reached a conclusion without technical support, persistent prodding, and relentless editing by my never-fazed research assistant, Melissa X. Stevens. My gratitude goes warmly to them all.

Conflicts of Interest:
The author declares no conflict of interest.

1
The Wife's Lament is one of nine so-called elegies in the Exeter Book as edited by Anne L. Klinck (Klinck 1992) with The Wife's Lament on pp. 210-11, and by Bernard James Muir (Muir 2000) with the poem's text in vol. 1, pp. 328-29 and commentary in vol. 2, pp. 664-67. R. L. Leslie edits the poem with useful commentary (Leslie 1961, introduction pp. 3-12, text pp. 47-48, notes pp. 53-58). R.D. Fulk added The Wife's Lament to his updated edition of John C. Pope's Seven Old English Poems making it Eight Old English Poems (Pope and Fulk 2000), with text on pp. 39-40 and discussion on pp. 120-28. Peter S. Baker includes the poem in his Introduction to Old English (Peter S. Baker 2012, pp. 207-10). The main editions I use in this essay are those by Leslie and Klinck, though I make reference to the others listed here. (Other introductory textbooks that include editions of the poem are not listed.) 2 Her gender is revealed by the feminine adjectives geomorre in line 1 and minre sylfre in line 2. Because the poet uses terminology associated with the comitatus (warband), some early commentators decided that the speaker was meant to be a man, one going so far as to delete the first two lines containing the "misleadingly" gendered adjectives: see Bambas (1963). 3 The following scholars have argued that because the Wife occupies an earthen dwelling (see part three of this essay), she must be dead and a revenant or speaking from the grave: Lench (1970); Tripp (1972); Johnson (1983); Semple (1998), and most recently Deskis (2020), although Deskis makes it clear that the imagery "does not require us to read the speaker in this way", but it "allows us to do so" (p. 385, her emphasis). Leanne MacDonald (2015) claims that The Wife's Lament is a poem about a zombie". The idea that the Wife's location is a grave is further promoted by a misunderstanding of the word "under" in the phrase under actreo at lines 28 and 36. "Under" in this sense is a British usage, as in the place name Boughton-under-Blean mentioned by Chaucer in line 3 of "The Prologue of the Canon's Yeoman's Tale" in The Canterbury Tales (Chaucer 1986). This village is lower on the hillside than the Blean Forest, just as the Wife's cave is situated lower on the hillside than the oak tree. For a firm rejection of the dead-Wife reading, see BeritǺström (1999). In a different approach, A.N. Doane argues that the Wife is a disaffected pagan goddess (Doane 1966), an idea revived by Banishalmah and Mizher (2020), basing their argument on the poem in Richard Hamer's translation (Hamer 1970). 4 Complex scenarios that include two or more men have been proposed for the poem. In agreement with recent editors, I regard the terms freond and hlaford as designating a single man beloved by the speaker. I will refer to him as her "husband", though whether they are married or not is irrelevant to my argument. 5 Niles does not insist on his own preferred view: "The speaker of the poem thus emerges, in the end, as a figure to whom individual readers can have very different responses, all of them legitimate, depending on their personal world view and values" (Niles 2006, pp. 205-6; see also 202, "door open"). 6 While Niles prefers the possibility that the Wife curses her husband ("I am felicitous in the 'curse' reading" (Niles 2006, p. 208)), he does not insist on his own preferred view: "The speaker of the poem thus emerges, in the end, as a figure to whom individual readers can have very different responses, all of them legitimate, depending on their personal world view and values" (Niles 2006, pp. 205-6; see also 202, "door open"). 7 Klinck examines and evaluates the use of the word "elegies" (Klinck 1992, pp. 223-51). 8 See Aeneid 4: 393-96. I am using the edition by Clyde Pharr (Pharr 1964), but the numbers of the books and their lineation are standard. Niles has argued for Ovid's Heroides as an influence on the poem (Niles 2019, p. 170). It should be noted, however, that the Aeneid was a much-quoted text in early medieval England, as Michael Lapidge documents in The Anglo-Saxon Library (Lapidge 2005, especially pp. 188-90); compare Lapidge's far scanter statistics concerning Ovid (Lapidge 2005, pp. 183, 323). 9 This essay in homage to J.D. Niles imitates the structure of his essay to which it replies, including his sidebar essay. While the sidebar essay within each longer essay can stand alone, each functions as an important part of the argument as a whole. 10 The most spectacular rock-cut architecture is in India and Cappadocia, but extensive buildings cut into rock exist in Malta, Slovenia, and elsewhere. Less known to those outside Australia is the currently inhabited rock-cut town of Coober Pedy in South Australia, and no doubt many other such dwellings exist throughout the world. These go unremarked for several reasons: a small cave shelter is literally less easy to spot than an equivalent small building on the open ground; the use of such buildings can slide easily between functions (dwelling place, storage, wine cave, etc.), and as living spaces cave-dwellings are readily associated with the poor and dispossessed, outlaws, and the supernatural. From the archaeological point of view, caves showing prehistoric occupation or ritual (such as art) are traditionally of major interest, medieval ones less so. 11 In a study of the noun hearg (herg, herh in the Northumbrian dialect), the archaeologist Sarah Semple says, "The OE term hearg is interpreted variously as 'pagan temple', 'hilltop sanctuary' and even 'idol'. It is a rare survival in the English place-name record. When it can be identified, the place name is commonly considered to refer to a location of pre-Christian religious activity, specifically a pagan Anglo-Saxon temple" (Semple 2007, p. 364). The pagan Danes in Beowulf worship aet haergtrafum (line 175), glossed "heathen temple" in Klaeber's Beowulf (Fulk et al. 2008, p. 391; but see the note on p. 128). 12 This is not a new idea; Christian W. M. Grein proposed it over a century ago (Grein 1865, p. 422) and Wentersdorf (1981) has constructed a long and well-supported argument for this reading. 13 Niles later modified that interpretation: "Since the phrase eard niman is used elsewhere in Old English to mean 'to take up one's abode', there is reason to take the phrase in that same sense here. The Wife's abode is then best viewed as a pagan herh, whatever difficulties may be involved in reconstructing the other details of her story" (Niles 2019, p. 237). As will be seen below, however, he retains the hostile husband view. 14 Baker says, "Editors do not agree on the interpretation of this line. Herheard is often glossed 'dwelling in the woods', but a herh (the more standard spelling is hearh or hearg) is a pagan shrine or sanctuary. [ . . . ] This edition retains the manuscript reading herheard in its obvious sense; the verse should be translated 'take up residence in a pagan shrine'" (Peter S. Baker 2012, p. 208, n. 7). 15 Note, however, that the word heard (hard) can have positive meanings such as "firm", "staunch", or "resolute". 16 For support in this reading, Klinck (1992, p. 181) refers to Nora Kershaw (1922, no page numbers cited) andW. S. MacKie (1925, pp. 91-93). 17 Leslie argues that the adverb her can alliterate, giving an example, in order to justify his emendation of heard to eard (Leslie 1961, p. 54), but the fact remains that his change weakens the alliteration. 18 Fulk points out that the initial h of heard is necessary for the alliteration; then he emends that word: "The reading selected for this text is based on a different assumption, that heard is a corruption of heord, more commonly spelled heorod, the chiefly Mercian equivalent of West Saxon hired 'household'" (Pope and Fulk 2000, p. 125). Fulk's editing of the poem conforms in principle to that of John C. Pope for the original seven poems of the book, by normalizing the text in order to allow students ease of access. Therefore his version of The Wife's Lament will vary from editions following the original Exeter Book text more closely. Fulk explains his method on pages xi-xii. 19 For the alternation of g/h as in hearg and herh, see Campbell (1983, p. 446). Hilda Ellis Davidson proposed that herh-os on the right side of the Franks Casket means "pagan deity" (Davidson 1969, pp. 216-26). See also Sarah Semple's archaeological study of the word hearg in place names (Semple 2007). 20 Later Niles changes his mind about this word, now accepting that herheard refers to a pagan sanctuary (Niles 2019, pp. 236-37), and this leads him to some interesting speculations that will be examined in part three of this essay. But he does not change his mind about the "hard-" hearted husband. 21 Inhabited caves and dugouts are often reused and enhanced from earlier natural cave or rock-cut structures. Such domiciles are the focus of part three of this essay. 22 An alternative suggestion in the same vein is that the husband has been outlawed by his own people for a violent act, perhaps like that committed by Beowulf's father when he took refuge with Hrothgar (Beowulf, lines 459-72; see Hamer's introduction to his translation, Hamer 1970, p. 71). Such outlawry was typically of a temporary duration, which could explain the verb abidan (await) in the final line of The Wife's Lament. The Wife may be hoping for her husband to return after accomplishing either his mission or his outlawry-if he survives. 23 As Klinck says, the phrase þissum londstede at line 16b refers to "a place that may or may not be the husband's own territory but is clearly alien to his wife" (Klinck 1992, p. 181). Her location in a friendless land not her own and the hostility toward her from her husband's people suggests the Wife's possible role as a "peaceweaver". (For an excellent history and analysis of this term, see Peter S. Baker (2013, pp. 103-26)). As she grew up the young woman would have been trained in diplomacy and hall ceremonies with a view to being married into a neighboring kin-group, thus becoming a sort of hostage in an attempt to keep peace. For an argument about misuse of the practice, see Alaric Hall (2006, pp. 81-87). The standard exogamous marriage frequently did not end well, as Beowulf himself points out at lines 2039-41. Though the violent outcomes when peace-weaving fails are the ones we hear about (as in the stories of Freawaru and Hildeburh in Beowulf ), sometimes the woman does succeed in accomplishing her task, as Wealhtheow does when she eases tensions in Heorot. For one recent argument among several disputing the old idea that Hrothulf will murder his cousin when the old king Hrothgar dies, see Osborn (2019). 24 "Oft" is an interesting word here. Has the poet personally observed or experienced such repeated "seizures"? Grief spasms may be a symptom of ASAD (Adult Separation Anxiety Disorder), a mental disorder only recently recognized by the DSM-5 (Diagnostic and Statistical Manual of Mental Disorders: DSM-5 2013). In some cases ASAD represents a persistence or recurrence of the childhood-onset type. If the Wife was married exogamously in her early teens, as was often the case, her current symptoms may represent a recurrence of anxiety about that earlier separation from family and friends. In any case, she is "disabled" by her violently emotional response to her mandated lonely situation. 25 One may be seized "by" a sudden panic attack, but we would not be likely to say, in the active voice, "panic gripped me". Thus the Wife speaking with today's usage would have to express herself more fully: "Very often here feelings about my lord's departure fiercely overwhelmed me". Such differences from ours in early English ideas about mental states are examined by, among others, Malcolm R. Godden (1985), distinguishing between the classical concept of a unified inner self and the vernacular idea of a mind separable from the inner self; Britt Mize (2008), on the mind as enclosure and container; and Leslie Lockett (2011), building on these discussions combined with ideas from modern psychiatry, such as Hinton and Hinton (2002), "Panic Disorder, Somatization, and the New Cross-Cultural Psychiatry". None of these scholars discusses the distress appearing as an outside agency (somatization) in some Old English poems, though Lockett touches on the concept of somatization itself (pp. 172-77). The Wife's panic attacks lead to her reflective contrast between her lonely situation "here" and the happier one of the couple living above ground. Throughout the poem it is the situation resulting from her husband's leaving that causes her anxiety (panic attacks) and makes her increasingly furious. 26 As has long been noted, the implicit chronology of The Wife's Lament is not matched by the order in which actions are mentioned in the poem. The word "curse", like "oath", is a multivalent term. Note that this mini-essay and its title allude to John D. Niles' mini-essay and its title, "Toward an Anthropological Theory of Cursing" in Old English Enigmatic Poems and the Play of the Text (Niles 2006, p. 187). 28 I believe that this sentence is intentionally ambiguous as the Wanderer (now a wise man) moves toward the ambiguous fastness at the poem's conclusion. The meaning of fastness includes both the fortress ruled by the [L]ord he has longed for and a secure situation of "stability", thereby combining the secular and spiritual longing into a single image of the Great Hall of God (Martin Luther's "Mighty Fortress of our Lord"). This doubling of secular and religious meanings builds up toward the end of The Wanderer, adding spiritual depth to the poem. There is nothing like it in The Wife's Lament with its focus on passionate love-longing. 29 The explanation in The Stanford Encyclopedia of Philosophy (Braun 2017) begins: "An indexical is, roughly speaking, a linguistic expression whose reference can shift from context to context. For example, the indexical 'you' may refer to one person in one context and to another person in another context. Other paradigmatic examples of words that function as indexicals are I, here, today, yesterday, he, she, and that. Two speakers who utter a single sentence that contains an indexical may be saying different things". 30 The suggested scenario where the husband in The Wife's Lament is constrained by an oath taken in the past adds temporal interest to the poem as it places the speaking Wife conceptually between the two "times" of her husband's oath: his "then" in making it, her "now" in lamenting the result for her (her loneliness), and his "later" (but soon, presumably) enactment of it, what he has gone over the seas to do. 31 Agamben (2011) with the discussion following that publication is an example of this interest. 32 Technically, a promissory oath is sworn publicly to an authority whereas a vow may be a personal promise made in private, although the "marriage vow", sworn to one's partner, is made publicly. It is doubtful that poets writing in Old English made such careful distinctions in their use of the related words ad (oath), beot (pledge), and treow (troth or vow). In the poetry all three terms appear to refer to a witnessed and irrevocable personal guarantee, so the oath-taker should consider carefully where it might lead before speaking such a promise in public (as in beot spriceð, in the passage quoted below). 33 Reference to lof as a worthy aim for a warrior is a structurally interesting element in Beowulf, coming at the beginning of the poem (lof-daedum, line 24), the middle (lof, line 1536), and the very end (lof-georn, line 3182). 34 See Magennis (1996, p. 119). 35 For an early argument that "the situation of the poem is that the wife of a lord has been made a prisoner in an oak grove at the instigation of her husband" because he has become hostile toward her, see Greenfield (1989b, p. 149). 36 This "noble custom" is specific to the warband. R.D. Fulk says of the word indryhten, "This adjective is probably formed from in-dryhtu, the prefix indicating a quality inherent in a noble dryht" (Pope and Fulk 2000, p. 193); in his glossary under dryhten he defines dryht as "army or host" (Pope and Fulk 2000, p. 176). 37 In a thoughtful discussion in Heroic Identity in Beowulf Scott Gwara (2008, pp. 173-77), questions the standard view that Hengest broke the oath that was meant to quell violence at Finnsburh. He makes a clever argument that is well worth consideration. 38 R. D. Fulk explains the element collenas the "pp [past participle] of a lost verb *cwellan, swell, spring up, grow big" (Pope and Fulk 2000, p. 173). 39 Andy Orchard traces a series of other words in The Wanderer that have both secular and Christian meanings (Orchard 2002, pp. 1-26). He finds that the ambiguity creates a structural development from the passive wanderer (eardstapa) waiting for ar (help, favor, grace, mercy, etc.) to the contemplator actively seeking it at line 115 (here translated "grace"). Orchard does not comment on the irony of these two stances as they proceed "from the selfish and worldly preoccupation with a lost past life to a selfless and unworldly concern for a timeless and future afterlife" (p. 12). But he does caution the reader that the ambiguities in the opening lines are not completely, as Greenfield argues, "artfully resolved at the poem's close" (p. 8, quoting Greenfield 1989a, p. 134). Even the phrase "Father in heaven" (faeder on heofonum, line 115), Orchard says, is "far from unambiguous", and he then lists a series of "key words and phrases" in the final lines that have both secular and Christian meanings, sometimes having secular meanings earlier within the poem itself: beorn, eorl, mid elne, treowe, ar, even frofre (Orchard 2002, pp. 6-7). The one word this fine close reader omits from this list is bot at line 113. In note 18 he translates this word "remedy", without comment. 40 The meanings of torn at line 112 and bote at line 113 are significant here. In their glossaries under torn, Leslie has "grief;" Bliss and Dunning have "resentment", then "anger on p. 59; Klinck has "remedy, amelioration". In their glossaries under bot, Leslie has "remedy"; Bliss and Dunning have "remedy": Klinck has "remedy, amelioration". For an acceptable meaning I would translate lines 112b-113: "[a warrior] must never too quickly reveal his anger from his breast unless he first knows the remedy" (i.e., how to ameliorate the situation, how to swing it his way, probably through violence). In this sentence the poet is not telling this (imagined) warrior to "stand down" or to "make peace", but to work out how to get payback most effectively and at least cost. 41 This is primarily Bjork's interpretation in "Sundor aet Rune", with which others concur. It should be noted that the Old English word bot meaning "remedy" occurs mainly in a medical sense, but so does our modern word "remedy" that is similarly used in other contexts. 42 Stanley claims that this half-line is "a difficult intrusion", and he then offers some suggestions made by others: "Gollancz (The Exeter Book 1895, p. 303) has 'unless he experience adversity', which uses adversity in a sense of 'opposition'. My 'contrariety' is nearer to the 'enmity' of faehþe witan, but no 'feud' is mentioned, and, if interpreted literally, 'feud, enmity' would divert from the paternal teaching. Shippey's 'unless he knows he has an enemy' (Shippey 1976, p. 51) and DOE's 'unless he perceives enmity' (s.v. fǣhþ, sense 1. 'feud, state of feuding, enmity, hostility; hostile act') would be diversions from the teaching of the poem" (Stanley 2018, p. 286). 43 Leonard Neidorf translates this precept: "A wise man seldom enjoys himself without worrying; just as a fool rarely mixes enjoyment with concern about his future, unless he knows he has an enemy [nefne he faehþe wite]. A sensible man must be careful with his words, and think things over in his heart, not be loud and noisy" (Neidorf 2021, p. 45). "Such advice is inherently aristocratic", says Neidorf, and arguing for a secular environment for the poem, he suggests that the speaker is imagined "as neither a monk nor a personified abstraction", as others have maintained, "but as an elderly Anglo-Saxon aristocrat endeavoring to persuade his ambitious son that piety is compatible with prosperity" (Neidorf 2021, p. 34 (abstract)). Whatever way one translates this precept, the sequence of ideas may refer to a cautionary formula to be heeded by a warrior when there is faehþe in the air. Following that significant word, the wisfaest haele of "Precepts" (line 57b) may remind one of the man snottor on mode ("wise in his heart") who sits apart in secret thought in line 111 of The Wanderer, or of the silent, pondering husband in The Wife's Lament.

44
This is my translation of the Rad stanza in M. Halsall (1981, p. 86). 45 Dennis M. Kratz (1980) argues that the peculiar final scene of the poem is intended as mockery. 46 The alternative proposal mentioned above, that the husband has already performed the mordor and is exiled "over the waters" because of that deed, does not fit the context so well as the idea that he is oath-bound to take action abroad from his homeland. 47 See further Sinan Oruc (2018). 48 The Wife may not have had a servant as devoted as those who looked after Guthlac, especially the monk Beccel who came to live with him during his last days (Colgrave [1956(Colgrave [ ] 2007, but she makes no mention of any deprivation except the comfort of her husband and friends. 49 Niles amplifies his idea of a pre-Christian setting in God's Exiles (Niles 2019, pp. 236-37). 50 For reuse of structures in the early English landscape, see, among others, Williams (1997, pp. 1-32); Semple (2013, passim); and Hartmann (2019, pp. 230-64). In "Barrow Agency: Reading Landscape in Felix's Vita Guthlaci", Jan-Peer Hartmann (2022) argues that the Vita Guthlaci illustrates how literature may transform perceptions of the landscape through the audience's expectation of genre. Felix initially describes Guthlac and his Fenland (a real person in a real location) partly on the basis of hagiographies believed to be known to the saint and the Vita's audience, but he brings in new elements as the text progresses. The Guthlac who results is no triumphant colonizer of a primeval wilderness; instead, he both influences and is shaped by the landscape in a symbiotic process of becoming. As an assemblage of variously interacting natural and cultural forces, Guthlac's Fenland is a physical and mental construct, as is the Wife's herheard according to this essay's argument, and this location also participates in the "process of becoming" that shapes the Wife in the poem. 51 Unlike the dwelling called a souterrain, dug down into the ground as the name implies and known mainly in Celtic areas of Britain (not in the poet's early medieval England), the rock-cut cave houses clustered in the Midlands are carved horizontally into the vertical face of a cliff. The fact that Battles and Garner do not consider this type of building along with the others that they examine demonstrates how new the cliff-face rock-cut building is to the typology of earthen dwellings. Moreover, attention is only now being directed to dating these buildings, so even if these two scholars had known about them, it is unlikely they would have considered them within an early medieval context. 52 Edward B. Irving, Jr., imagining the little fox with her kits as representing a woman in a cave, claims that the riddle is "about people driven to act like animals" and "dragging their children, hands clapped over screaming mouths, out of the way of some marauder" (Irving 1994, p. 204). I consider the subject of this riddle to be a realistically imagined animal whose actions under duress bear analogy with human ones, in particular those attacked by Vikings in the ninth-century. Hardly the panic-stricken mother that Irving imagines, the vixen first delays, then effects the "remedy" (bot) with a counter-attack. Harried by the enemy she describes as an unwanted "guest" in line 10 and identifies as a waelhwelp (a terrier who "goes to ground" after prey) in line 23, she takes the precaution of digging an exit route from her burrow to lead her kits to safety, and only then does she return to set viciously upon her attacker. 53 For dates see Colgrave's introduction to Felix's Life of Saint Guthlac (Colgrave [1956(Colgrave [ ] 2007. 54 On the basis of the poet's extensive description, Niles is able to identify with some certainty the way we are meant to understand the nature of the fire coming out of that human-built "cave" (see Niles 2012, pp. 25-36). Niles' description of what Beowulf sees as he looks into the dragon's abode is a model of the information that can be retrieved from a poem about an imagined place, a site based on, he says, "a megalithic chambered tomb (or long barrow) of the kind built in many parts of Europe during what we now know as the Neolithic era, with this building activity culminating around the years 2300-2200 BC" (27). 55 Aeneas objects strongly to Dido's interpretation of their night together in the cave (Aeneid 4: 338-39), and he points out that it is not his choice to go to Italy (Aeneid 4: 361). He never stops loving her deeply (see Aeneid 4: 393-95). 56 According to a Historic Environment Records (HER) database reported by the City Archaeologist Scott Lomax at Nottingham City (Lomax 2021). 57 For presentation here, the print has been cropped in order to emphasize the cave entrance at the end of the winding road. The inscription along the top reads, in its entirety, "View of Blackston Cave, River Severn & Ld. Herberts house near Bewdley, Sep. 23." .