‘Exeter Riddle 4’ and Two Other Bell Riddles

Abstract

In early medieval England, the ringing of bells served a wide range of functions. This article argues that a detailed understanding of bells and their everyday use can help to explain the intricate narratives and elaborate language of three Latin and Old English riddles. The first, the late antique ‘Riddle 80’ by Symphosius, describes a dinner bell and the idle chatter of a drunken party. The second, Tatwine of Canterbury’s ‘Riddle 7’, presents a funeral bell hanging in a tower as a deposed emperor who is hung and beaten. The third, the well-known ‘Exeter Riddle 4’, presents an everyday monastic bell as if it were an obedient monk, and it casts their relationship as interdependent and symbiotic. The bell solution has been heavily disputed. However, when the riddle is read alongside the rich context of monastic culture, and with careful attention to linguistic detail, this solution is confirmed.

The bells of early medieval England came in all shapes and sizes, and they were rung for a myriad of purposes, both religious and secular. Bells ranged in size from small, clinking tintinnabula and wooden tabulae to large, clanging circle bells. They announced all kinds of events, both prosaic and extraordinary, from mealtimes or religious ceremonies to the death of an abbess or abbot, and the sighting of an invading force. We should not think of the quintessential early medieval bell as an enormous circle bell ringing out sonorously from large belfries and cathedral towers. In their study of medieval church bells, John Arnold and Caroline Goodson write that:

The origins and development of bells in the church took diverse and circuitous routes […]. At worst, the bell-tower lurches precariously back in time, and all its associated uses and meanings impose themselves upon every past mention of bells (or of what are assumed to be bells). ¹

As one would expect given their profound cultural importance, there are many references to bells in late antique and early medieval literature. Riddles were one of the most important and prestigious literary genres of early medieval England, and so it should be no surprise that bells featured in several riddles too. The trend begins with Symphosius’ ‘Riddle 80’, one of the 100 Latin riddles of this anonymous late antique author. His collection, probably written in North Africa at some point between the third and sixth century, was extremely influential on the early medieval riddle tradition. ² The eighth-century riddle collection composed by Tatwine, archbishop of Canterbury, also includes a rather idiosyncratic bell riddle, ‘Riddle 7’, in his collection of forty Latin riddles. ³ In both cases, as we will see, an understanding of the differing functions of the bells described in these riddles can provide us with us a deep and nuanced understanding of the artistry of these riddles. Many scholars would also include a third bell riddle, ‘Exeter Riddle 4’, which is one of approximately ninety-five Old English riddles in the late tenth-century Exeter Book. ⁴ However, these riddles do not include answers. As a consequence, the solution to this complex riddle has been highly contested ever since Franz Dietrich suggested a bell in 1859, along with the alternative solution of millstone. ⁵ Since then, many solutions have been suggested, including lock, ⁶ handmill, ⁷ pen, ⁸ bucket, ⁹ dog, ¹⁰ the Devil, ¹¹ plough-team, ¹² sword, ¹³ and the sun. ¹⁴ Craig Williamson, writing in 1977, considered the riddle to be unsolved. ¹⁵ More recently, Patrick Murphy has robustly defended the bell solution, as a ‘symmetrical metaphor of a bell as a creature roused in the same way that a reluctant riser responds to the ringing of a bell’. ¹⁶ His reading reminds us that riddles of this kind do not merely describe objects, but the social relationships that underpin them too. Despite Murphy’s work, the bell solution is not yet secure enough to be considered definitive. This article intends to redraw the lines of scholarly opinion firmly on the side of the bell. To do so, we must appreciate how closely aligned the riddle’s complex and enigmatic language is with the rich literary and social context of bells and bell-ringing.

Dinner bells and Symphosius’ ‘Riddle 80’

Most bells in Ancient Rome were small bells or gatherings of bells, which were known in Latin as tintinnabula. References to these bells are common in classical Latin literature, and they attest to a wide variety of functions and associations. They were used as protective pendants, ¹⁷ phallic windchimes, ¹⁸ ceremonial bells, ¹⁹ doorbells, ²⁰ and livestock-bells, ²¹ amongst other things. ²² Bells may also have been used as alarms or for announcing the watches, although these functions were normally reserved for trumpets. ²³ Many bells had a connection with time: announcing bathing times in public baths, ²⁴ the time for household slaves and employees to get out of bed, ²⁵ funerals, ²⁶ the opening and closing of markets, ²⁷ and the change of courses at banquets. ²⁸ It is probably the last of these functions that Symphosius has in mind in his ‘Riddle 80’, which is solved as tintinnabulum (‘small bell’).

Aere rigens curvo patulum componor in orbem.

Mobilis est intus linguae crepitantis imago.

Non resono positus, motus quam saepe resulto. ²⁹

As Leary notes, the context of this riddle immediately before those on food and drink, along with the riddles’ symposium context, suggest the bell for the changing of courses at dinner. ³⁰ The first line explains the shape and material of the bell. According to Leary, the association of bronze with stiffness and hardness juxtaposes with the image of the bell’s curvature, ³¹ which creates a minor opening paradox. But the reference to aes is as much an etymological as a physical clue since instruments of all kinds, including bells, were often referred to metonymically and in the plural using this word. Writers would often refer to bells or cymbals by pairing aes with the word tinnitus (‘jangling’). For example, both Donatus and Augustine use the phrase ‘tinnitus aeris’ to describe onomatopoeia. ³² The stock phrases ‘tinnitus aeris’, ‘tinnitus aere’ and ‘tinnitus aeni’ can also be found in the work of Seneca, ³³ Ovid, ³⁴ Catullus, ³⁵ and Ausonius, ³⁶ among others. It seems likely that the first word of the riddle is designed to evoke both the jangling sound of the bell and the first two syllables of the solution. The first and second lines describe the bell’s rim as if it were a wide-open mouth in orbem, and the clapper as a talkative tongue. These clues are not merely visual, but associative too. The propensity of small bells to jingle constantly, rather than ring out deeply, led to their association in classical literature with sciolistic and empty chatter. As Leary explains, ‘while able to “talk”, bells could not speak rationally and misogynistic comparisons were therefore made between female garrulity and the sound of bells’. ³⁷ For example, when describing gossiping women, Juvenal’s ‘Satura 6’ tells us: ‘verborum tanta cadit vis | tot pariter pelves, tot tintinnabula dicas | pulsari’. ³⁸ A related proverb, originating with Plautus, has it that ‘numquam […] temere tinniit tintinnabulum; nisi quis illud tractat aut movet, mutum est, tacet’, ³⁹ with the meaning that a gossip needs an audience. ⁴⁰ The idea of a chattering bell also fits well with the performative context of Symphosius’ riddles – the party atmosphere of the classical symposium. Not only does the jangle of the bell announce the different courses, but it also mimics the drunken chatter of the guests. We are reminded of Symphosius’ remarks on drunken blabbermouths in the prologue to his riddles, when he describes those guests who tell riddles as ‘verbosa cohors studio sermonis inepti’. ⁴¹ A similar idea is expressed in the first two lines of Symphosius’ ‘Riddle 70’, which describes how a water clock regulates speech and silence.

Lex bona dicendi, lex sum quoque dura tacendi,

Ius avidae linguae, finis sine fine loquendi. ⁴²

Although this riddle is about how a clock regulates the time granted for an orator to speak in a legal or administrative setting, ⁴³ it also alludes to the idea that Saturnalia is a festival on which the courts are ‘silent’ and only idle chatter is permitted. ⁴⁴ In a similar way, the chattering bell regulates the time that is allocated to human chattering at the symposium, reminding us such time to tell riddles is limited. This may explain the acrostic AMMON(E)O (‘I remind or warn’) that is formed by the first and last word of each line, albeit requiring the addition of one letter. In such roundabout ways, the riddle reveals an unexpected connection between the chattering bell and the chattering riddle-teller.

Tower bells and Tatwine’s ‘Riddle 7’

Many of the functions of classical bells continued well into the medieval period. For example, the Old English law code known as the ‘Hundredgemot’ mentions three forms of sound that can be admitted as legal testimony: ‘hryðeres belle, hundes hoppe, blæshorn’. ⁴⁵ The implication is that oxen and guard-dogs in tenth-century England often wore bells, just as their classical Mediterranean forebears did. Likewise, the Roman use of bells at funerals continued into the Middle Ages in the ringing of the bell at impending death (i.e. the passing bell), at the news of the death of a holy person (i.e. the death knell), ⁴⁶ and the ringing of various kinds of bells at funerals (i.e. the lych bell). ⁴⁷ Just like their classical ancestors, most bells of the period ‘would presumably have clanged or tinkled, rather than tolled sonorously across a distance’. ⁴⁸ Although larger, mounted bells and more sophisticated casting techniques became increasingly widespread in the eighth, ninth and tenth centuries, the large church bell hung in a belfry was a comparative rarity in England until the twelfth and thirteenth centuries. ⁴⁹ This is not to say that belfries did not exist, but only that the bells were not typically of the scale found from the end of the eleventh century onwards. However, various church buildings, known as tower naves, were built in pre-Conquest England, many of which would have housed small and medium-sized hanging bells. Built in imitation of imperial architecture on the continent, tower naves were typically high-status chapels, often with a mortuary function, which required a bell. ⁵⁰ The earliest-known examples date from the eighth and early ninth century. ⁵¹ The first example was constructed by Wilfrid at Hexham shortly after regaining his Northumbrian see in 706. As one would expect from Wilfrid, he consciously adopted a Roman style of architecture, in direct contrast to Irish-influenced Lindisfarne; the building likely functioned as a high-status mortuary chapel. ⁵² The two other bell towers known to have been built in the eighth century – at York and Winchester – were also used for funerary purposes. ⁵³ Tower-building reached its pre-Conquest apotheosis during the tenth-century Benedictine Reform, when it became the mark of a prosperous monastery, such as Glastonbury, Abingdon, Thorney, Sherborne, Worcester or Winchester. ⁵⁴ The Colloquies of Ælfric Bata, written at the end of the tenth century, lists a cloccarium or bellhus among those things that a good monastic church should have. ⁵⁵ Many of these Reform-era buildings were also closely associated with royal authority. For example, Byrhtferth of Ramsey describes King Edgar as standing high up in the tower at Winchester Old Minster as the sacristan rings the bell for Vespers, in his Vita Oswaldi. ⁵⁶ A related type of bell tower developed at the end of the tenth century, when it became fashionable for lords to construct non-monastic tower naves, often situated independently from ecclesiastical buildings. For example, Wulfstan of York’s ‘Geþyncðo’ mentions a bellhus among the property requirements of a thegn. ⁵⁷ Such edifices were typically intended as extensions to high-status residences, and their design borrowed heavily from the iconographical language of imperial prestige and ecclesiastical authority. ⁵⁸

Tatwine’s ‘Riddle 7’, written in the early eighth century, describes a hanging bell, possibly an early tower bell of some kind. ⁵⁹ Tatwine clearly does not have the small, jangly bell of Symphosius in mind, and he gives his riddle the unusual title, De tintinno, perhaps in the mistaken belief that tintinnabulum is the diminutive form of a noun, tintinnus. ⁶⁰ Tatwine would likely have been aware of the early Northumbrian bell towers; he would certainly have seen continental examples in his visit to Rome to receive the pallium. ⁶¹ Intriguingly, his riddle appears to refer to a tower bell’s mortuary function as well as its association with imperial power. Sandwiched between riddles on the essential materials of the scriptorium (letters, pen, parchment) and the church (altar, cross, lectern), it describes its subject using a narrative about an emperor’s misfortune.

Olim dictabar proprio sub nomine caesar,

Optabantque meum proceres iam cernere vultum.

Nunc aliter versor superis suspensus in auris,

Et caesus cogor late persolvere planctum

Cursibus haut tardis cum ad luctum turba recurrit.

Mordeo mordentem labris mox dentibus absque. ⁶²

The riddle relies upon two parallel narratives. In one, an emperor is deposed and tortured, before taking his revenge. In the other, each time the bell rings, it is as if it is being beaten and its cries are from the pain. The bell is described in the opening line as an emperor. The suggestion by Franz Buecheler that this is a reference to the melting down of bronze statues or coins of pagan emperors to be recast as bells is somewhat plausible. ⁶³ However, Tatwine’s primary inspiration is most likely a pseudo-etymological pun between the verb caedere (‘to strike’) and the familia Caesar, ⁶⁴ a name that was once glorious and renowned by proceres. The metaphor also alludes to the high status of the hanging bell, and perhaps to its associations with Rome and imperial power. The emperor is then overthrown, hanged and tortured, which leads to the pun’s return in line four, when the caesar (the ‘striker’) is himself caesus (‘struck’). Likewise, the bell is raised and hung up and then struck.

Depicting a hanging object as if it is being tortured or executed is not unique to Tatwine’s riddles. For example, the oil lamp of ‘Lorsch Riddle 10’ denies that it is a criminal, before admitting that ‘in laqueo reus ut fur pendeo longo’. ⁶⁵ Alcuin uses a similar idea in the Disputatio Pippini, when Pippin announces cryptically, ‘Nunquam bene, nisi suspendantur in aere’, and Alcuin answers him with the equally enigmatic ‘Audivi mortuos multa loquentes’. ⁶⁶ The remark is designed to mislead the reader that the dead hanging on the gallows can somehow talk, but the real meaning is the funeral bell that ‘speaks’ for the dead. However, Tatwine’s tortured bell remains alive, which allows for another piece of wordplay, this time on the dual sense of planctum as lamentation (a meaning derived from the idea of beating the breast) and the simple noise of striking. The emperor cries out in his pain and the bell rings out as if it wails. The bell’s luctus attracts the masses very quickly (cursibus haut tardis). In the literal narrative, these crowds appear to be gathering happily to witness the torture of a pagan emperor; ⁶⁷ in the metaphorical narrative, they are thronging to the sounds of a funeral, with the implication that the bell is a lych bell. The final line borrows heavily from Symphosius’ ‘Riddle 44’ on the onion, which proclaims ‘mordeo mordentem […] dentesque non habeo ullos’. ⁶⁸ Just like his predecessor, Tatwine is imagining the bell as a mouth, and it is very likely that he had Symphosius’ tongue-like clapper in mind. The biting refers to how the clapper and the lip of the bell’s rim strike – or ‘bite’ – each other. At the same time, he implies that the deposed emperor has had his teeth bashed in, yet he is somehow taking his toothless revenge on those who earlier harmed him.

Monastic bells during the Benedictine Reform

Although early medieval bells played an important role in announcing deaths, the two most prominent uses of bells were to mark the canonical hours and to call a congregation to church. The origins of the bell as a marker of prayer has traditionally been ascribed to Paulinus, the fourth-century Bishop of Nola, on the basis of the late Latin word nola (‘mass bell’). ⁶⁹ A separate tradition ascribes this to Pope Sabinian at the turn of the seventh century. ⁷⁰ But, given the various ways in which bells were already used for announcing time in the pre-Christian world, it seems likely that bells were used to mark Christian prayer and ceremony much earlier than these legends would have it. The bell was particularly important to monastics, who aimed to live a life of constant devotion, in which every act, however mundane, was an act of prayer. Thus, bells were increasingly used to mark out not only the canonical hours, but also each moment of the heavily regulated monastic day – thus a bell would ring whenever monks were expected to begin a new task.

The increasingly complex lives of the monasteries and churches of tenth and eleventh England required a considerable number of different bells, often in double figures. In the early eleventh century, the minster at Sherburn-in-Elmet, a moderately-sized but wealthy institution, owned ‘.iiii. handbellan, ond .vi. hangende bellan’. ⁷¹ Similarly, the Exeter inventory drawn up by Leofric in the years after the Conquest records that ‘ond þær næron ær buton .vii. uphangene bella, ond nu þær sind .xvi. uphangene ond .xii. handbella’. ⁷² The mid-tenth-century monastic rule, the Regularis concordia, is a testament to the variety of roles that bells performed, mentioning three different types of monastic bell. The first is the tabula, a wooden panel that was probably adopted from the practice of Roman basilicae. ⁷³ According to the Regularis concordia, the tabula is to be struck whenever there is manual labour to be carried out during the day, ⁷⁴ for Saturday maundy, ⁷⁵ and when death seems imminent. ⁷⁶ The second kind of bell, the tintinnabulum, was a small hand-bell that was usually rung for some time. ⁷⁷ It is often associated with night-time prayer in the Regularis concordia, marking out periods of transition such as the ordered procession to church, and so the tinkling of tintinnabula must have been a common one in the early hours of the evening and morning. ⁷⁸ But the most frequent bell referred to in the Regularis concordia is the ambiguous signum, which was rung for daytime church gatherings, ⁷⁹ prayer, ⁸⁰ confession, ⁸¹ mealtimes and other events in the monastic day, as well as for seasonal customs, such as the ringing of all the bells at Nocturns during the Christmas period. ⁸² The signum was typically rung by the sacristan or their secular equivalent, the ostiary, whose identities were strongly bound up with such timekeeping. For example, in the Monasteriales indicia, a handbook of monastic sign-language, the symbol for the cyricweard (i.e. the sacristan or ostiary) is a hand-movement ‘swylce he wille ane hangigende bellan teon’. ⁸³ Their role included timing the bell correctly, as other insular and continental monastic texts attest. For example, the ninth-century Institutio Canonicorum Concilii Aquisgranensis includes a chapter on why and how canons should observe the hours most diligently, explaining that ‘custodes praeterea ecclesiae harum horarum distinctiones bene norant, ut scilicet signa certis temporibus pulsent …’. ⁸⁴ Similarly, in his pastoral directions to the clerics of Sherborne minster, Ælfric requires that the ostiary ‘sceal mid bellan bicnigan þa tida’. ⁸⁵

Since the bell marked out the rhythms of the monastic day, it also became a symbol of monastic obedience and submission. Just as a monk was bound to follow rule and custom unquestioningly, so they were bound to obey the bell unquestioningly too. For example, in a discussion on the weekly maundy each Saturday, the Regularis concordia stresses that ‘nullus quippiam quamuis parum sua ac quasi propria adinventione afere presumat […] sed ab universis […] tempore opportuno consuete peragatur’. ⁸⁶ Obedience did not simply require that one should drop everything at the sound of the bell, but also that one should not anticipate it. For example, the Regularis concordia describes the period of study or prayer between Prime and Terce thus:

Tunc facto signo eant et se diurnalibus induant calceamentis; nullus enim hoc debet praesumere antequam illud audiatur signum, exceptis ministris neque tunc ab aliquo intermittatur sine licentia ne praesumptione temeraria oboedientiae meritum lugubriter obnubilet. ⁸⁷

Although the monastic schedule did grant some space for individual expressions of piety, these were also ultimately regulated by the bell. Thus, although the Regularis concordia allocates time for personal prayer after the first call to bed, the final call is to be obeyed without exception.

Si quis vero post haec, devotionis suae forte fervore, his diutius incumbere volverit, agat quidem haec; sed audito signo aeditui quo resides ad egrediendum vocat nil moretur.’ ⁸⁸

In such a way, the bell calls out to the ordinary monk and the exceptionally devoted alike. Its ringing was the bottom line, and its power to stir monks to action was legendary in the traditions of monastic literature. For example, in Bede’s Vita Cuthberti, a paralysed monk is sent to Lindesfarne, where he is laid out with relics next to his feet. He lies there all night, until ‘at ubi consuetum in monasterio nocturnae orationis signum insonuit, excitatus sonitu resedit ipse’. ⁸⁹ The bell calls him to rise from bodily paralysis, just as it calls his brothers to rise from sleep. At the same time, monks were not expected to be passive automatons, but rather to obey wilfully the signal on each occasion. ⁹⁰ Another recurrent literary motif was the idea that the bell rouses the righteous from sleep just as Christ’s teaching rouses them from eternal death. For example, an anonymous Old English homily, ‘Assmann XIV’ ⁹¹ warns its audience that ‘deofol us læreð slæpnesse and sent us on slæwðe, þæt we ne magon þone beorhtan beacn þære bellan gehyran’. ⁹² The homilist then tells them that God ‘us læreð wæccan and sent us on leohte heortan, þæt he wolde, þæt we oft cyrican sohtan’. ⁹³ In this way, each ring was a call to rise from the dangerous inertia that imperilled the soul – the spiritual consequences of disobeying the bell were very serious indeed.

‘Exeter Riddle 4’

Only twelve lines long, ‘Exeter Riddle 4’ uses a compelling variety of enigmatic words and phrases to disguise its subject in a way that is at once harmonious and baffling.

Ic sceal þragbysig þegne minum,

hringum hæfted, hyran georne,

min bed brecan, breahtme cyþan

þæt me halswriþan hlaford sealde.

Oft mec slæpwerigne secg oðþe meowle

gretan eode; ic him gromheortum

winterceald oncweþe. Wearm lim

gebundenne bæg hwilum bersteð;

se þeah biþ on þonce þegne minum,

medwisum men, me þæt sylfe,

þær wiht wite, ond wordum min

on sped mæge spel gesecgan. ⁹⁴

The first and probably the most difficult clue is found in the opening half-line – the compound adjective þragbysig, which is not found anywhere else in Old English corpus. Its apparent novelty and its unusual position in respect of the governing pronoun would both seem to indicate especial significance as a clue. Any reading of the riddle must therefore begin by asking what þragbysig means, as well as explaining its unusual syntax. This is no easy task. The meaning of the þragbysig is not immediately clear, not merely because it is a hapax legomenon, but also because of the difficulty of interpreting the þrag- element. Unfortunately, Bosworth & Toller’s dictionary definition of þrag as ‘a time, season’ has misled several scholars. ⁹⁵ It is clearly referring to the phrase ða þrage and ða tide in ‘Blickling Homily 11’, ⁹⁶ which is tempora vel momenta in the Vulgate and the times or the seasons in the King James Bible. ⁹⁷ But none of the sixty-six instances (fifty-one in poetry and thirteen in prose) of þrag refer explicitly to the seasons. ⁹⁸ Although this would not refute seasonal solutions such as Shannon Ferri Cochran’s plough-team or Ann Harleman’s bucket of water, it would certainly weaken them significantly, since both scholars lean on this meaning considerably. ⁹⁹ Importantly, þrag always designates a time-period (either as a noun or adverbially), rather than a particular time or season, or the idea of abstract time. ¹⁰⁰ Although its near synonym, tid, can refer either to dates and events in time (a feast day, a canonical hour, a mark on a sundial) or to periods of time (e.g. for a time, sometimes, from time to time), þrag only ever refers to the latter. It is frequently used as a dative of time (þrage or þragum). ¹⁰¹ The dative singular has the sense of ‘for a time.’ Often this is a relatively long period of time, ¹⁰² but such a meaning is relative, and so the ninth-century Corpus Glossary simply glosses it with the Latin word interim (‘for a while’). ¹⁰³ The plural adverbial form þragum is less common, occurring nine times in the corpus, always in poetry. On each occasion, it expresses repeated periods of duration, with the sense of ‘for several periods of time.’ ¹⁰⁴ When used nominally, þrag means ‘for a while’, and never ‘at some particular time.’ For example, in ‘Guthlac B’ it refers not to the moment of the loss of a lord, but to the aftermath, since one grieves in the wake of death, rather than only at the moment of death.

Ellen biþ selast þam þe oftost sceal

dreogan dryhtenbealu, deope behycgan

þroht þeodengedal, þonne seo þrag cymeð,

wefen wyrdstafum. ¹⁰⁵

Old English poets often describe periods of hardship, and this might mislead us into thinking that þrag has intrinsically negative connotations. In Juliana, it refers twice to the period of the Devil’s punishment, ¹⁰⁶ in the Old English Boethius to the period of Boethius’ imprisonment, ¹⁰⁷ and in Beowulf the compound earfoðþrag (‘a period of hardship’) ¹⁰⁸ refers to the twelve years of Heorot’s oppression. However, þrag can also refer to benign or positive times, such as the period during which a patient must inhale vapour from a kettle ¹⁰⁹ and the length of Christ’s time on earth. ¹¹⁰ Thus, if we accept Melanie Heyworth’s contention that þragbysig ‘introduces a sense of the potential for negativity in this riddle’ ¹¹¹ then we must also accept that the converse is also true.

The -bysig element of þragbysig is much easier to understand. The adjective bysig, the noun bisigu and the verb bisgian all refer to the process of occupying oneself mentally, whether positively in productive work or negatively in cares and anxieties. ¹¹² We should not discount the importance of the -bysig element, with its sense of mental intentionality. When considered alongside the evidence above, this tells us that the compound þragbysig is unlikely to mean ‘seasonally busy’, ‘busy with (’abstract’) time’, or ‘busy at a difficult time’. It also seems incorrect to read þragbysig simply as a synonym for the ubiquitous and more metrically flexible hwilum (‘at times’, ‘sometimes’, ‘frequently’). Since the noun is headed to an adjective, the meaning of þragbysig is unlikely to be a copulative (‘at a time and busy’) or appositional (‘at a busy time-period’) compound. If it is endocentric, it means ‘busy at a time’, ‘busy for some time’, or perhaps even ‘busy all the time’. Alternatively, if it is exocentric, then it represents something more than the sum of these two words, for example ‘preoccupied’, ‘hard at work’, or ‘obedient’.

The unusual position of þragbysig after noun and verb raises another important question about its meaning: is the object always þragbysig, or only when it sceal… hyran…brecan…cyþan? The linguistic evidence would point to the latter. The use of postnominal adjectives (i.e. adjectives that follow their noun) to qualify a pronoun is occasionally found in OE verse. ¹¹³ Although postnominals with ic are a comparative rarity, the form does occur several times in the Exeter Riddles, perhaps under the influence of similar, appositive constructions in Latin riddles. Thus, we find the following:

1. 1. Þonne ic astige strong ¹¹⁴

2. 2. Þonne ic winnende | holmmægne biþeaht, hrusan styrge ¹¹⁵

3. 3. Þe ic lifgende lond reafige | ond æfter deaþe dryhtum þeowige ¹¹⁶

4. 4. Nelle ic unbunden ænigum hyran | nymþe searosæled ¹¹⁷

5. 5. Þonne ic hnitan sceal hringum gyrded … ¹¹⁸

6. 6. [Ic reorde] hwilum mæwes song | þær ic glado sitte ¹¹⁹

The exact sense of the adjective or participle is ambiguous in several of the above examples, and I do not wish to touch on the debate as to whether these postnominals are attributive (e.g. ‘I, strong, rise’), predicative (e.g. ‘strong, I rise’), or appositional (e.g. ‘I rise, strong’). ¹²⁰ But I would note that in every instance here, the adjective or participle describes the governing pronoun as it was at the time of the main verb. Thus, the wind is strong when it rises and it fights when it covers up the ocean, the ox ravages the land when living, the bow is disobedient when it is unbound, the key is bound by rings when in use and the jay sits cheerfully when singing. On this evidence, it seems likely that the subject of ‘Exeter Riddle 4’ is only þragbysig when it sceal… hyran…brecan…cyþan. When we recognise that this is only the case when it ‘obeys its servant’, ‘breaks the bed’ and ‘announces loudly’, then certain aspects of the riddle’s prosopopoeia come into focus. A bell is busy and hard at work when it is rung; at other times, it sits idle. This makes an interesting contrast with the monk, who is always expected to be busy and obedient, but especially so when the bell is rung.

Servitude and obedience are constant themes throughout the riddle, but particularly in the first four lines. In lines 1b and 2b, the riddle plays with the ostensibly paradoxical idea that a master might be obedient to their servant (þegne minum…hyran georne). Scholars have explained this in various ways. For example, Donne, Cochran and Stewart all suggest that the riddle creature is an object used by a person who is themselves a servant, ¹²¹ and Steward and Heyworth interpret the word þegn ambiguously as both ‘servant’ and ‘lord’. ¹²² But the ‘obedient to my servant’ paradox is most easily reconciled when we understand it in terms of the symbiotic relationship between bell and monk – each is obedient to the other. The monk is servant to the bell, to whose sound he must hyran georne without question; the bell is a servant of the sacristan or ostiary who rings it. Furthermore, as Murphy points out, the dual meaning of hyran as ‘to hear’ and ‘to obey’ is particularly suitable for the bell solution. ¹²³ The apparent paradox also chimes with other expressions of monastic obedience in texts of this period. As Katherine O’Brien O’Keeffe has shown, these frequently went beyond simple gestures of behavioural compliance to create a complex ‘drama of freedom and obedience’ that complicated the relationship between superior and subject. ¹²⁴ Given this literary context, it is entirely fitting that ‘Riddle 4’ should also blur the distinction between ‘servant’ and ‘lord’.

The theme of servitude is continued in line 2a, hringum hæfted, which is the first of three punning references to rings. The word hæftan is extremely suggestive of servitude. It can mean ‘to bind’ but also ‘to capture or enslave’, from which the word hæft (‘prisoner’, ‘slave’) is derived. This may allude to an aspect of the bell’s binding or mounting: the loophole (hring) of a hanging bell is fastened (hæfted) to its mounting, and the loophole of a hand bell is tied to a leather strap. But, as Murphy notes, the phrase need not refer only to the bell’s fastening, but also its ringing. ¹²⁵ Certainly, this sense of hring is attested by the existence of compounds such as belhring and nonhring with a similar meaning in the OE Longer Rule of Chrodegang. ¹²⁶ We could say that a bell’s entire life is bound with ringing – it has no other existence. Similarly, the monk, in their commitment to total obedience, is bound tightly to the ringing of the bell. Line 3a explains that the bell must min bed brecan, which Murphy links to the early morning ringing that woke monks from their beds before Nocturns and Matins. ¹²⁷ This phrase also seems to express a subtle monastic irony. The bell spends most of its time unused and ‘in its bed’. When it is roused from its sleep and rung by its master, monks are also roused from the dangerous sleep of lethargy. Another important clue is found in in line 3a – the bell must cyþan breahtme. Cyþan typically refers to a verbal act, and so the idea is that the bell is, in a sense, speaking. Yet it does so breahtme, a word that can refer to a loud noise, but which can also mean ‘a glimmer’, and thus ‘in a flash’, as attested by its glossing of atomus in the ‘Cleopatra Glossaries’. ¹²⁸ Both senses are appropriate to the loud, abrupt clanging of a bell. It announces ‘that’, or quite possibly ‘because’ (þæt), its monastic master has given it a halswriþa. ¹²⁹ Tupper thought that this word suggested a leather bell-harness, ¹³⁰ but it more likely refers to the large ring (the loophole or ‘crown’) at the top of a handbell or hanging bell. If one imagines that the bell is a person, then we see the sloping upper edges as the shoulders and the portion above this as the neck – we still refer to these features as the shoulders and neck today – and so the crown is literally the ‘neck-ring’. The social and religious connotations of halswriþa also fit well with the passage’s theme of servitude and obedience. It could suggest a benevolent lord bestowing a ‘necklace’ on a favoured retainer, or equally it might imply a punishment collar, a symbol of slavery inherited from Rome. ¹³¹ Finally, it also plays upon the bell’s similitude to the monk, who rises ‘in a flash’ to proclaim his servitude to God, which is signified physically through another kind of ring that sits above the neck – the monastic tonsure. All these connotations of halswriþa are therefore consistent with the bell interpretation.

From line four, the theme of obedience dissipates for a time and other aspects of monastic bellringing come into focus. The bell is described as slæpwerige, which continues to play upon the dual idea of the bell awoken from its ‘rest’ and the monk awoken from lethargy. The person who oft…greteþ the bell on each occasion is the sacristan or ostiary. This might be secg oðþe meowle, depending on whether the monastery houses monks or nuns. The idea of ‘touching’ a sleepy bell suggests the common monastic practice of rousing slæpwerige monastics quietly by touching them if they oversleep. ¹³² Gretan has a secondary meaning of ‘to play an instrument’, ¹³³ and it is possible that the riddle plays upon this sense of the word too. The idea of an early morning wake-up call is continued with the phrase winterceald. Not only does this describe the cold-to-the-touch iron of the bell, but also the monk whom it wakes. The bell then oncweþe, repeating the conceit that the bell speaks. It does so to the weary sleepers, who might, in their unwillingness to rise, easily be described as gromheorte, ¹³⁴ particularly if we see riddle as a kind of self-deprecating, monastic inside joke. More humour can be found in the idea of a wearm lim that gebundenne b[æ]g hwilum bersteð, which several scholars have understood as a not-so-subtle innuendo. ¹³⁵ Tupper and Murphy both interpret the wearm lim to be either the ringer’s arm or the clapper. ¹³⁶ Tupper also explains the gebundenne b[æ]g as the band that secures the clapper, although the bell’s circular rim seems more likely. The success of this reading depends on how we understand the verb berstan. It is always used intransitively or with a dative indirect object in Old English except on this occasion, ¹³⁷ but a transitive sense of bersten and to burst is common in Middle and Modern English. It usually denotes breakage or rupturing, but it seems reasonable under the circumstances to understand it as denoting a falling, crashing, or smashing, with a meaning of ‘at times, the clapper crashes against the bell’s rim’. Such a reading seems appropriate when considered alongside examples of berstan elsewhere in Old English verse to denote loud, violent events such as crashing thunderclouds ¹³⁸ and breaking waves. ¹³⁹ The riddle uses berstan to express the violent suddenness of the bell’s ringing – a literal and allegorical wake-up call to the weary monastic.

The final four lines shift the riddle’s focus back to the opening lines’ themes of interdependence and symbiosis, inviting us to contemplate how the monk and the bell are dependent on each other. Again, this is described as a relationship between the speaker and their servant (þegn). Both are depicted as equally unwise and ignorant. The monk is the bell’s servant because he is bound in obedience to its call, and he is medwisum because he requires a bell to wake him from their literal and figurative slumber. At the same time, the bell that the monk obeys can neither wiht wite nor wordum min | on sped…spel gesecgan because it can neither ring without its servant ringing it nor speak intelligibly. Not only does this play with the idea that the monk serves a master who is entirely ignorant and unthinking, but it also recalls the classical trope, as found in Symphosius’ ‘Riddle 80’, that a bell’s ringing is merely empty chatter. Despite this, the message of the bell is agreeable (on þonc) to the monk, who is thankful for its sound – however gromheort he might be – because it calls him from the dangerous spiritual inertia that threatens the soul.

Conclusion

‘Exeter Riddle 4’ is one of a small group of socially aware riddles that play upon the relationship between human and bell. All three riddles use prosopopoeia to imagine the bells as creatures with mouths that speak in curious ways, although they are otherwise quite dissimilar in both the language that they use and the kinds of bells that they describe. The tradition begins with Symphosius’ urbane and highly self-aware ‘Riddle 80’, which uses its subject to satirise the verbose symposium environment of the classical riddle, gently mocking the very riddle form itself by comparing it to the dinner-bell’s empty patter, as well as reminding us that the time for such games is limited. The complexity of religious life in early medieval England required various types of bells with various functions, and so it should not be a surprise that this complexity is reflected in the riddles that came after Symphosius. Tatwine’s ‘Riddle 7’ personifies a mortuary bell as a deposed emperor, a choice that reflects the high status of such devices as well as the apparent connection between the words caedere and Caesar. Although the riddle reflects the social use of such bells to announce deaths and funerals, the relationship between human and bell is largely etymological, rather than being based on any deeper level of similitude. ‘Exeter Riddle 4’, on the other hand, is saturated in the discourse and ideology of everyday monastic bellringing. It dramatizes the symbiotic relationship between bell and monk, which is presented in terms of the master-servant relationship that is so common to the riddling tradition. By focusing on the obedience of both the bell who is rung and of the monk who must respond to the ringing, it reminds the reader that the bell’s call was not to be answered blindly or unwillingly, but rather consciously, wilfully, and dutifully. In such a way, ‘Exeter Riddle 4’ is as much about the experience of living by the bell as it is about the bell itself. It is perhaps the most thoughtful consideration of the relationship between bell and monk that survives from early medieval England.