“Once Upon a Time There Was a Saint…”: Re-evaluating Folklore in Anglo-Latin Hagiography

Abstract

This article examines methods for identifying folklore in hagiography. Using hagiographical materials from eleventh and twelfth century England, it critiques the current trend of equating folklore motifs with oral transmission and argues in favour of a “performer-centred” understanding of folklore and hagiographical composition.

The Rehabilitation of Folklore

Hagiographic texts describing the lives of centuries-old saints are notoriously problematic sources for modern-day scholars. Their highly derivative nature is an often-cited cause for condemnation as the same tedious and clichéd stories occur time after time with predictable regularity. The standard biographical structure of birth, life, death and posthumous miracles tend to be fleshed out with routine topoi and fantastical digressions. For Hippolyte Delehaye, an early historian of hagiography, these texts represent “certain audacious fabrications, products of lying and ambition [which] have for long misled over-credulous minds and unwary critics” (Delehaye 1998, 78). More recent historians have been equally damning: “the authors of these Vitae were writing chronological nonsense; and what is more, avoidable nonsense… They did not care: or anyway, they did not bother” (Campbell 1986, 225). This sense of frustration arises from the expectation that hagiography ought to be historical, that these texts ought to be historically accurate. But to believe this rather misses the point.

For a start, developments since the “linguistic turn” have discredited the idea that historians can ever enjoy unmediated access to past realities. We now appreciate that the historical past as represented in medieval histories and chronicles is problematical and cannot be accepted without reservation. Even if we were to strip hagiography of its seemingly fictitious elements and retain just the bare “historical facts” the genre still does not lend itself to an uncritical reading; the historical past as presented was not without authorial mediation. If hagiography cannot be read for its historical content, a more useful approach is to reconstruct its historical context: not what the texts say about the past, but what they disclose about their present. For hagiography to be of service to cultural historians, a holistic approach must be taken which apportions equal weight to both the apparently factual and the fictitious. Furthermore, in order to locate hagiographic texts within their original socio-cultural context it is important to consider both their function and reception.

There is broad agreement that the primary purpose of hagiography was edificatory, to provide enlightenment through the power of example. Yet the contexts in which the texts were disseminated and how they were received are often overlooked. Saints’ legends were transmitted in a wide variety of ways and, correspondingly, there were a received by a wide variety of audiences. For example, a prose vita or miracula may have been used for private devotion, read aloud in a paraliturgical sefng such as a monastic refectory or chapter, or marked up for use in the Office on the saint’s feast day. Metrical vitae were also composed, as were sets of antiphons and responsories for use in the liturgy. Lengthy texts were often abridged and incorporated into vast compilations, such as John of Tynemouth’s Sanctilogium Angliae, Walliae, Scotiae et Hiberniae. Vitae even provided the basis for vernacular sermons and popular verses. A single saint’s vita therefore may have been made to serve many uses. As a result, hagiographic texts had to operate on several levels: their liturgical deployment called for clear passages of biblical and exegetical commentary to expound spiritual truths; their auditory performance necessitated a certain literary quality; and their sermonic application demanded features guaranteed to appeal to a vernacular audience.

Recognition of the need for universal appeal has resulted in a reappraisal of the historical importance of hagiography. Numbered alongside exempla-collections, sermons and penitential manuals, hagiography has been perceived to have lain at the cultural interface between the Latinate culture of the clerical elites and the vernacular culture of the laity. These texts were instruments designed by the clergy to correct abuses and compel the laity to comply with orthodox practices (Gurevich 1988, 2). The idea of the clergy and the laity in opposition, competing over belief and orthopraxis was first articulated by Jacques Le Goff who propounded the “two cultures” view of medieval society. This is a sociological model which presumes that social grouping was the main determinant of belief. Medieval society (and religious belief) was imagined as polarised along several axes; elite and popular, clerical and lay, learned and unlearned, written and oral, official and folkloric. These opposing categories of society were locked in continuous conflict: “Clerical culture opposed folklore not only out of conscious and deliberate hostility but equally out of incomprehension” (Le Goff 1980, 158). But while the “learned” culture unremifngly sought the suppression of folkloric culture, clerical elites also saw the benefit in appropriating choice elements to frame orthodox Christian teaching (Picard 1989, 373). Theologians writing pastoral literature were inclined “to stoop to the mental level of [the] parishioners” and “resort[ed] to familiar images, confining themselves to subjects within the mental horizon of their flock” (Gurevich 1988, 3, 12). Folklore thus became the vehicle by which people were brought to orthodoxy as practised at the ecclesiastical centre.

This model has received considerable criticism. Carl Watkins has recently presented a persuasive case against using fragments found in normative texts to reconstruct popular religion, arguing that their innate conservatism and generality undermines their value as sources for local particularism (Watkins 2007, 8). Moreover, the dividing line between official and unofficial cultures was far from clear. The capacity for individuals to belong to more than one social group made things far more ambiguous. In England, the geographical focus for this study, monastic recruits often came from aristocratic families; their early education would not have differed from that dispensed upon their thegnly or baronial brothers. To propose a distinction between monk and warrior is unjustified. The distinction between parish priest and layman may be equally unwarranted. The priests of manorial churches were most likely selected from among the lords’ tenants or by fellow villagers. They are categorised alongside villeins and bordars in Domesday Book and their livings are indistinguishable from standard peasant holdings (Blair 2005, 492). Although Gregorian reformers called for a stricter form of sacerdotalism, a celibate, educated and non-hereditary clergy in step with diocesan agendas was slow to emerge (Watkins 2007, 7). On the eve of the twelfth century most priests were part of and on a par with the local community and would have participated in the shared culture of the village, including the circulation of local folklore.

Recent historiography has emphasised the overlap between clerical and lay culture. John Blair is convinced that closer and more integrated links existed between the two communities, and has argued that the folkloric elements found in hagiography are indicative not of clerical appropriation, but of participation: “Legends may have passed backwards and forwards between laity, who circulated them orally, and monastic writers, who gave them coherence and moral purpose” (Blair 2002, 478-79). A similar conclusion has been reached by John McNamara in relation to the construction of Bede’s Historia Ecclesiastica, and by Julia Smith with regard to Breton hagiography (McNamara 1994; Smith 1990). The presence of folklore in hagiography therefore has become equated with the oral transmission of traditional material and evidence for interaction between lay and clergy. Hence folklore has come to be regarded in a positive light and rehabilitated as a bona fide object of historical enquiry. Proving the mutual participation of clergy and laity in the construction of saints’ legends has the potential to be extremely useful. It may reveal something about the nature of the interaction between laity and clergy, those areas where their world views coincided and even shed light on storytelling processes.

This approach provides a corrective to both Delehaye and advocates of the “two cultures” model, but it is not without its own deficiencies. It requires refinement not least in how folkloric elements might be identified. Some have followed Delehaye and have drawn a distinction between the conventionally hagiographic and more colourful material which they assume must owe to a lively oral tradition: “The more idiosyncratic the material… the more likely that it derives either from a more specific hagiographical tradition or from local vernacular culture” (Blair 2002, 479). Other methodologies are even less empirical. In her article on Breton hagiography, Julia Smith claims that the Vita Corentini “retains many features characteristic of oral traditions,” but omits to declare explicitly what these are (Smith 1990, 327). This article therefore aims to redress this insufficiency and set out a methodology by which folkloric elements in hagiographic texts can be more securely identified.

Finding Folklore

Scholars working on Celtic hagiography have long been alert to the crossover between saints’ vitae and folktales. Studies by Elissa Henken on Welsh hagiography and Dorothy Bray’s counterpart study on Irish saints indicate significant areas of overlap (Henken 1991; Bray 1992). Their conclusions are founded on the identification of parallels between hagiographical elements and folk motifs as listed in Stith Thompson’s Motif-Index of Folk-Literature, which, at six volumes is considered the most consummate index of its kind (Thompson 1955-58). Such an exercise may also be attempted with Anglo-Latin hagiography; vitae may be pared down to their constituent narrative elements and identified one at a time in Thompson’s Motif-Index as the legend of the Dorset saint Juthwara demonstrates. [1]

Juthwara was a young noble virgin, who, after the death of her father, was left in her stepmother’s care. Her wicked stepmother conspired to rid herself of her stepchildren; Juthwara’s brother was exiled and Juthwara treacherously slandered. Pretending to be administering a remedy for her pallor and lassitude, the stepmother placed two pieces of fresh cheese on Juthwara’s chest and instructed her to walk to the church. Meanwhile, her brother returned eager to disprove the rumour which had begun to circulate that Juthwara was pregnant. He joined the crowd gathering outside the church. When Juthwara emerged she denied the allegations, but the rabble, believing the melting cheese seeping through her clothes was breast milk, beheaded her. The decapitated girl rose, picked up her head and walked back to the church. Where her head had fallen a spring erupted and a tree grew. When cross-referenced against the Motif-Index this story might be told as follows: a cruel stepmother (S31) orders her stepdaughter to be killed (S322.4.2) and exiles a boy (S322.4). The girl is slandered as an adulteress (K2112) and the brother flogs his unchaste sister to death (Q458.2.1). The deception is broken or the disenchantment brought about by decapitation (D711). The girl carries her head under her arm (F511.0.4) and spring rises where the head was cut off (D925.1.2), while a tree miraculously grows (D2157.4) from the blood of the slain girl (E631).

The legend of Kenelm lends itself to a similar analysis, but while the motifs constituting the legend of Juthwara bear liele relation to one another and possess no thematic logic, the motifs in the Vita et miracula S. Kenelmi form narrative clusters, still recognisable to modern readers ([BHL 4641] Love 1996, 49-89). Instead of breaking the legend down into motifs, its most elementary form, it may be analysed by tale type. These are tales which possess a particular constellation of folk motifs (Aarne and Thompson 1961; Thompson 1977; Uther 2004; “ATU” hereafter denotes international tale types in Uther 2004). After his father’s death, the seven-year-old Kenelm inherited the kingdom of Mercia. He had two sisters, Burgenhild who loved her brother dearly, and Cwoenthryth who was intensely jealous and plotted his death. After failing to kill Kenelm with poison, she promised his tutor a portion of the kingdom if he were to succeed. One night Kenelm dreamt he was standing at the top of a tall tree when it was suddenly cut down and fell to the ground with a crash. He imagined that he became a little white bird and flew away. Kenelm’s nurse explained that this dream foretold his death. Shortly afterwards, his tutor took him hunting. Kenelm grew exhausted and while he slept the tutor dug a shallow grave. Upon awakening, Kenelm resisted his murder and planted his staff in the ground which grew into a great tree. He was dragged to a neighbouring valley and decapitated. He caught his head in his hands and a white dove flew away. Where the body was buried the grass grew lush and green and a cow pasturing there produced miraculous quantities of milk. The murder was revealed when a dove delivered a letter containing all the details to the Pope in Rome. When Kenelm was finally disinterred, a spring erupted in his grave and his body was taken to Winchcombe. Cwoenthryth tried to stop the translation by chanting psalm 108 (109) backwards, but her eyes fell out and when she died soon afterwards her body resisted burial in consecrated ground.

The motifs in this story constitute clusters which display a thematic logic recognisably familiar as belonging to specific folk tales. John Blair has drawn attention to its similarity to the tale “The Juniper Tree” (ATU 720) (Blair 2002, 481-2). In the Grimm Brothers’ version a young boy, killed by his stepmother, is accidentally eaten by his father. His sister gathers up the bones and buries them beneath a juniper tree, near his mother’s grave. A beautiful bird flies away and its sweet song tells the story of his murder. The motifs of the wicked stepmother, the good sister, the boy who becomes a bird and a bird which reveals the truth are found in both stories. Catherine Cubie has suggested that Kenelm’s dream contains elements from ATU 328, better known as “Jack and the Beanstalk” (Cubitt 2006, 192). Other tale types can be identified. The jealous stepmother, the attempted poisoning, the enlisting of another, and the woods as the location chosen for the murder, are all found in ATU 709 or “Snow White.” Although we know that Kenelm is a martyr and that the story cannot possibly end well, when the first attempt fails we still hope the tutor will turn into the nice woodsman and allow Kenelm to run off into the forest. The legend of Kenelm fuses together an assortment of fairy tales. Although we now recognise them as “Snow White” or “Jack and the Beanstalk,” these titles are new and the plots themselves relatively recent. It is only since they have been written down that their forms have become fixed. Prior to that these motifs were fluid, capable of being arranged and rearranged. Andrew Lang remarked that fairy tales are like “a kaleidoscope: the incidents are the bits of coloured glass. Shaken, they fall into a variety of attractive forms; some forms are fitter than others, survive more powerfully, and are more widespread” (quoted in Fentress and Wickham 1992, 62). Could it be that that in these vitae we are catching a glimpse of the motifs before they become crystallised in fairy-tale form?

Although attractive, this argument is profoundly ahistorical and sheds no light as to how hagiographers came by these motifs nor what meanings they held in eleventh-century England. In fact this approach can only be considered admissible if historians are prepared to accept a flawed definition of folklore. Recently Catherine Cubitt declared her interest in “popular oral stories.” She defined these as “folktales, stories which circulated orally and were transmitted not primarily through learned and written sources” (Cubitt 2006, 189), which feels intuitively correct because we bring a priori assumptions to bear on the compound term “folktale.” “Folk” implies a contraction of “common folk” or “common people” which, in itself, connotes oral transmission (orality being the type of transmission characteristic of the “common folk”) (Ziolkowski 2007, 51). Hence a folktale is presumed to be a collaborative or group creation by common people which was orally transmitted. This reasoning however is defective. The concept of “common people” is inapplicable to medieval society; the popular culture of folktales, folk dances and folk songs belonged as much to the social elites as to the masses (ibid. 52). Moreover, it would be folly for medieval historians, for whom philology and the identification of textual relationships are major interests, to discount the possibility of textual transmission. It is conceivable that even the most idiosyncratic hagiographic motifs circulated in solely textual forms.

This highlights another mistake historians have made in identifying hagiographical folktales. The selection of the hagiographic episodes to look up in Thompson’s Motif-Index tends to be decided on the basis of a motif’s presumed folkloric content. Catherine Cubitt, for instance, concludes her study of Abbo of Fleury’s Passio Sancti Eadmundi with the observation that as there were “no hagiographical models for these elements … they must derive from a lively oral tradition”(Cubitt 2006, 197). Oral stories or folktales are defined by default: they are what remains once the biblical, hagiographical and classical motifs have been identified. This rigid categorical thinking presumes no interplay between these genres. In fact, folk motifs surface in classical works, Apocryphal texts and canonical Scripture. [2] Moreover, this was a two-way process: motifs of classical or Biblical extraction were reciprocally absorbed into folktales. Thus a motif which appears to be conventionally hagiographic may have wended its way into a folktale and ended up entering the vita through folklore. [3] By assuming a hagiographical origin we run the risk of overlooking truly folkloric items.

Current historical methods for defining and identifying folklore in hagiography are clearly inadequate. At the heart of the problem is the assumption that motif indices are arbiters of folklore. This, however, indicates a profound misunderstanding of the point of motif indices and the research objectives of their compilers working in the Finnish comparative tradition. Adherents of the Historic-Geographic method (by which the comparative method is also known) applied a philological approach to the analysis of folkloric “texts.” This involved collecting the many variants of certain tales and statistically extrapolating a hypothesised archetype from the evidence of these variant texts. Their aim lay in determining the origin and distribution of individual tales. Put simply, historic-geographic folklorists were essentially interested in variants and in establishing the extent of variation involved. Historians, however, use motif indices to play a different game and search for similarities, points of convergence. In so doing, they are using the indices in ways they are not intended to be used and for analyses for which they are ill-suited. Medievalists searching for bona fide folklore will not find it playing snap with a motif index.

Thus we return once again to our original point of departure; how does one identify hagiographical episodes which may have surfaced from oral tradition? In her study of Breton saints, Julia Smith stated that Latin clerical culture “bore the heavy imprint of oral ways of thinking” (Smith 1990, 311), while John Blair commented on stories which “follow the rules of oral transmission” (Blair 2002, 487). From this we might venture that not only does oral tradition leave its traces in written texts, but that it conforms to a series of rules and conventions. But what exactly were these rules and what do these traces look like on the written page?

These were the questions which stimulated Milman Parry and Alfred Lord to devise their Oral-Formulaic theory (Lord 1960; Parry 1971). They were interested in proving the oral basis of Homeric epic and developed their theory after closely observing the processes of oral composition and transmission in contemporary Slavic epic poetry. Extrapolating from the examination of live practices they outlined a series of formulaic and thematic criteria which they thought were indicative of extemporised composition. These formulas and themes, they argued, were the essential tools of the illiterate singer. They then extended this to claim that any text exhibiting a high percentage of formulas and formulaic systems possessed an “oral” provenance (Lord 1960, 131; Foley 1997, 616).

This theory has both its apologists and detractors. Applauded for its innovative approach and wide application, it is also criticised for providing only analogies not proof, and for not taking into account differences between languages or the retention of formulaic idiom in written forms (Bäuml 1984; Foley 1997, 618). Its application to medieval Latin hagiography is particularly challenging because rigidly-metered epic poetry composed and transmitted in the vernacular bears few similarities to prose saints’ Lives skilfully crafted in Latin for use in a (para)liturgical setting.

An even-weightier concern which truly limits the theory’s applicability is the fact it does not allow for personal artistry. It cannot cater for variation and hinges on the notion that “formulas” facilitate extempore composition. Thus for a text to be “oral” it must present similar formulaic structures and its written form should inevitably contain a considerable degree of uniformity to other “oral” texts. Over the past forty years, however, new thinking has challenged the view that stability is a criterion of oral tradition. In fact, this approach takes a contrary perspective by proposing that variants are the natural by-products of the oral transmission process. Variation is borne out of the interaction between tradition, teller and audience (Dégh 1995, 175). In this scenario, the storyteller earns his reprieve; no longer the passive vehicle through whom the tradition passes, he is an active figure, an artist who selects, shapes and transmits his tale according to his aesthetic sensibilities. The audience also play an active role in censoring elements of the artistic performance, reining in more radical departures from the received tradition, while occasionally being the driving force behind the formation of new variants (ibid., 202). Performance is therefore an act of negotiation; individual poetic licence interacting with the traditions sanctioned by the community. The result is the birth of new variants, sub-tales and motifs.

Adherents of this performer-centred approach to folklore stipulate the need for fieldwork and the close observation of live-storytelling situations. [4] This ethnographical aspect makes its application to medieval hagiography problematic; it is hard enough to observe the dissemination of a tale in the present but impossible to do so in the past. We certainly should not place our trust in the authenticity of the storytelling descriptions sketched out by our hagiographers. [5] Nevertheless, there are several aspects to this thesis which medieval historians would do well to take away and apply. The first of these is the idea of variation. Variants point to places where communities shared items of folklore and, as this sharing would have been face-to-face, variants are the vestiges of oral transmission. Variation may be diachronic and/or spatial. The second concept to take away is the idea of conduits, the chain of individuals who received and transmitted the legend. Their individual proclivities had an enormous impact upon the course a tale took and the written shape it acquired. While it may be impossible to identify the tradition bearers individually, it is conceivable that comparisons across a textual hagiographical tradition may shed light on the conduit through which the narrative passed.

In attempting to visualise the act of storytelling we have come along way from the existing historiographical methodology which seeks to classify, quantify and compare folkloric items in hagiography. It now only remains to put this approach to the test and to apply this method to the Anglo-Latin hagiographical corpus.

Oral History in Latin Hagiography

Ecgwine was a seventh-century bishop of Worcester and founder of the abbey of Evesham whose hagiography contains traces of diachronic variation. The first Vita S. Ecgwini [BHL 2432] was composed by Byrhtferth of Ramsey in the early eleventh century. His task, initiated at the request of the Evesham monks, was particularly challenging for very little was known about this bishop. Byrhtferth had at his disposal only a handful of facts that he had been able to glean from charters in the abbey’s archive and the stories that the monks had told (Byrhtferth of Ramsey 2009, 208). Michael Lapidge, the modern editor of the text, argues that Byrhtferth supplemented this sparsity “in imaginative, but idiosyncratic ways” by drawing on allegory and hagiographical topoi (ibid., lxxxiv). Furthermore, Lapidge dismisses the two episodes which particularly concern us as merely products of Byrhtferth’s “fertile imagination” (ibid., xciii). The first of these episodes recounts how Ecgwine, having shackled himself and thrown the key into the river Avon before journeying to Rome, recovered the key inside the stomach of a salmon in the Tiber (ibid., 230-32). The second episode relates how the site of Evesham was revealed through a vision of the Virgin Mary who appeared to a swineherd as he searched for his pig (ibid., 244-248). Both of these episodes have classical antecedents, hagiographical parallels and entries in Thompson’s Motif-Index of Folk-Literature. [6] But it is by looking later, at their subsequent retellings, that we might better understand from where these episodes derived and what they may have meant to the local community.

The abbey’s foundation legend found in Byrhtferth’s Vita S. Ecgwini is a curious tale. Upon receiving a grant of land by King Æthelred, Ecgwine divided the land into four and apportioned a swineherd to each quadrant. One day, the sow which belonged to the swineherd named Eoves disappeared, only to reappear after several days with a litter of seven piglets in tow. This happened once again, only this time the piglets were all white, except for their ears and feet. [7] On the third occasion, Eoves went in search of the sow, which he found lying in a bed of thorns and suckling her piglets under the watchful gaze of the Virgin Mary. He reported his vision to the reeve who alerted Bishop Ecgwine and the bishop hastened barefoot to the forest. After witnessing the vision himself, the bishop ordered a monastery to be built on the selfsame spot.

Eighty years later, Byrhtferth’s vita was superseded by one written by Dominic of Evesham, sometime prior there ([BHL 2433] Lapidge 1978, 77-104). Dominic’s Vita S. Ecgwini was split into two books; the first comprised a revised edition of Byrhtferth’s vita and the second contained an account of Ecgwine’s more recent miracles. In the vita, Dominic modernised Byrhtferth’s archaic Latin, leaving out the abstruse allegory and numerological eccentricities. But besides adding a miracle story and substituting a more extensive charter and grant of privileges, Dominic retained Byrhtferth’s narrative structure. That is, in every respect bar the pigs. The legend of Evesham’s foundation was radically abridged. Eoves was transformed from a subulcus into a pastor and the legend of the absconding sow and her piglets was completely excised (ibid., 84-85). The rest of the story remained unchanged; upon seeing the apparition, Eoves notified Ecgwine who ventured into the forest to see for himself. Here Dominic concurs with Byrhtferth, the Virgin Mary was more brilliant that lilies, rosier than roses and smelled divine. It is likely that Dominic was working closely with a copy of the earlier vita. The part about the pigs was therefore a deliberate omission.

There are several reasons which might account for the rebranding. First, Dominic may have been keen to play up the parallels with the New Testament. In the chapter he reminds his readers that the angels revealed the news of Jesus’s birth to a group of shepherds. [8] This analogy would have lost its clout had Eoves remained a swineherd. Secondly, Ecgwine’s fame had spread. By the end of the eleventh century, his feast day was celebrated not only at Evesham and Worcester, but beyond the diocese of Worcester in East Anglia and south-west England (Wormald 1934, 97, 209, 251 and 265). His relics were even taken on a fund-raising tour of the country to generate revenue for the rebuilding of the abbey (Thomas of Marlborough 2003, 102-16). Dominic may have been responding to the demands of a wider audience. Confronted with the need to render Ecgwine a more universal figure, Dominic may have felt that stories which grounded the saint in the local area would have made him a less attractive intermediary to people in other parts of the country.

The story of Eoves and his sow, however, continued to hold currency. Within a decade of the publication of Dominic’s Vita S. Ecgwini a recension was produced, dubbed the “Digby-Gotha recension” by its modern editor (Lapidge 1979, 42-51). Intending to produce an abbreviation of Dominic’s text, the redactor retained his vocabulary and phrasing and rather than paraphrasing superfluous passages, he elected to omit them in their entirety. Hence, whatever differences exist between this text and Dominic’s were consciously undertaken. The reappearance of the sow and her successive litters of piglets is therefore of great interest and must represent a conscious effort to reintegrate this tale into the hagiography of Ecgwine.

This automatically raises the question how did the redactor come by this information? Michael Lapidge has reasonably suggested that he may have had access to a copy of Byrhtferth’s text (ibid., 40). Given that the redactor had no scruples in plagiarising Dominic, we might expect to find Byrhtferth’s text purloined to a similar degree. Yet the presence of several narrative inconsistencies and the lack of verbal and syntactical parallels mean we cannot prove that he had access to the earlier text. In fact, they rather suggest that he did not. The first point of departure is the absence of the names of the three other swineherds in the “Digby-Gotha recension,” although this may have been done for reasons of brevity. Of greater significance are the discrepancies in how the three litters are described. According to the earlier vita, the first litter consisted of “four and three piglets,” the second litter were all “white, except for their ears and feet” and (may have) numbered eight and the third litter contained nine piglets (Byrhtferth of Ramsey 2009, 246-48). The later “Digby-Gotha recension,” however, states that seven piglets were produced in both the first and second litters, while a “more numerous share” of piglets were born in the final litter (Lapidge 1979, 45). If the redactor had access to a copy of Byrhtferth’s vita these mistakes are inexplicable, yet are more pardonable if we regard these variations has having developed through oral transmission. The redactor was clearly telling the story in his own words, yet these are words which may have filtered down through the mouths of successive storytellers. As they each retold the tale subtle modifications crept in, numbers were mixed up and forgotten and colours missed out. Nevertheless, these were only fractional variations and the core of the story remained unchanged.

A second episode in the Vita S. Ecgwini exhibits signs of similar variation. In the first vita, the key to Ecgwine’s fetters, thrown into the Avon, was later found in the Tiber (Byrhtferth of Ramsey 2009, 230-2). But by 1125, when William of Malmesbury wrote his Gesta Pontificum Anglorum, the story had altered:

If we are to believe the old story, he [Ecgwine] once… shackled his feet and threw the keys into the river… He went to Rome in this state, and came back unscathed. But while he was crossing the strait between Gaul and England, a monster of a fish (so the story goes) leapt into the ship carrying the bishop… They gutted it, and found in its liver the keys, which when applied to the lock of the fetters freed the saint (William of Malmesbury 2007, vol. 1, 453).

The relocation of this miracle from the Tiber to the English Channel suggests that this tale had also entered into common parlance and folk tradition. Variants in the story of the key and the fish, like that of the sow and her piglets, were probably introduced during oral transmission. Over the years these stories would have circulated freely and widely, undergoing subtle modifications with every retelling.

Moreover, these stories indicate just how fluidly tales swung back and forth between oral and written expression; their every written reappearance was a snapshot of the latest twists and turns the tale had taken. Although we cannot claim to know with any authority how Byrhtferth came by these stories, whether he fell back on his classical education, improvised from his “fertile imagination,” or really did have his ear to the ground, we can see that these stories were picked up and passed on by word-of-mouth, becoming integral to the folklore of Ecgwine. [9]

Variation occurs diachronically and/or spatially. Carl W. von Sydow was the first to draw attention to the need to study regional idiosyncrasies in folklore. Adapting Darwinian principles to the study of folklore, he arrived at the concept of oikotype (or oicotype) to denote a tale which adapted by natural selection to a particular environment (von Sydow 1948; Clements 1997). He believed that the specific details of tales vary according to the surroundings in which they are told; the natural environment, economic structure, social system and cultural tradition (Honko 1980, 281). Although the responsibility for the composition of a tale lies ultimately with the individual performer, it is, at heart, a communal creation. The audience are participants in the construction and regulation of their lore and have the communal power to veto tales and stories which deviate too radically from the accepted norm. On the other hand, the social requirements of a community are subject to change. Communities are capable of taking, leaving or even inventing stories to satisfy their current needs (Geary 1994, Remensnyder 1995). The audience, therefore, may be a stimulus for modification as opposed to a source for stabilisation (Dégh 1995, 202). Finding evidence for oikotypification indicates that a particular tale has been adapted by a community for their use.

The legends of three West-Country saints, Juthwara of Sherborne, Sativola or Sidwell of Exeter and Urith of Chielehampton present a good example of oikotypification in practice. All three saints were the unsuspecting victims of a scheming stepmother who orchestrated their deaths. The earliest surviving account of Juthwara’s martyrdom appears in John of Tynemouth’s Sanctilogium. Probably abbreviated from an earlier vita, it recounts how Juthwara’s stepmother fooled her brother into believing Juthwara was pregnant thus inciting his anger and provoking him to decapitate her as she left the church (Horstman 1901, 98-99). A set of lectiones compiled by Bishop Grandisson in the 1330s contains the earliest extant vita of Sativola ([BHL 7488m] Grosjean 1935). Sativola’s death was similarly engineered by her stepmother. Having initially threatened and then bribed a group of haymakers, the stepmother sent Sativola to the meadow, whereupon the wicked servants cut off the girl’s head with their sickles (ibid., 364). Our knowledge of Urith of Chittlehampton derives solely from a Latin hymn serendipitously copied into a fifteenth-century notebook by a Glastonbury monk (Chanter 1914, 297-98). Although shorter than the vitae of Juthwara and Sativola, the hymn relates how Urith also met her death at the hands of harvesters on her stepmother’s orders. The parallels do not end there: Juthwara and Sativola were both said to have caught their heads with their hands and, in the places where the three girls were beheaded, springs were said to have erupted.

Finding three such similar stories within so localised an area inevitably leads to talk by some scholars of tales influencing others and the search for the definitive legend (Orme 1992, 171). Medieval commentators even tried to rationalise the similarities by claiming a sororal relationship. A twelfth-century Kalendar from Exeter refers to Juthwara as sororis sce. Sativola uirginis (Doble 1940, 17-18). In Juthwara’s fourteenth-century vita she is furnished with three sisters: Eadwara, Wilgitha and a certain “Sidewlla” (Horstman 1901, 99).

It is very possible, however, that the legends all had separate origins. In an eleventh-century catalogue of the relics housed in Exeter cathedral there is a reference to “the innocent virgin St. Sativola who, guiltless, was killed by her father’s pasture-man, and Almighty God afterwards revealed a multitude of miracles at her tomb” (Conner 1993, 186-87). The first reference to Juthwara also dates to the eleventh century. In Goscelin of Saint-Bertin’s Vita S. Wulfsige (written c. 1078-1080) she makes a brief cameo appearance: “Once upon a time she was beheaded by her brother… and after her head had been cut off, her mutilated body is said to have run with it and with both hands to have put it back on the neck from which it had fallen” (Love 2005, 115). Despite the fact that neither of these early references is particularly expansive, there is clearly no overlap between the legends at this stage. There are no signs of the treacherous stepmothers who feature so prominently in the later legends. It seems likely that, at some later point, an external narrative became subsumed into two different tales about two local heroines, creating two new oikotypes.

The assimilation of a new tale to an extant legend has been described by Lauri Honko as either “tradition-morphological” or “interior” adaptation (Honko 1980, 283). He explains that “…to have admission to the repertoires of narrators, new tales and themes must pass both collective and individual filters of preference… to become active, tradition must be adapted to a previously existing system of communication [otherwise] a new element might be simply rejected as controversial” (Honko 1980, 283). New traditions need to be linked to “tradition dominants,” such as historical figures, heroes, supernatural beings and other role models recurrent in the folklore of a social group or region. If an incoming story requires reshaping and recasting with familiar characters, the community reaches for local tradition dominants. Juthwara, Sativola and presumably Urith were the tradition dominants for their local communities, the mainstays around which new traditions were adapted. The slight differences in the articulation of the new traditions—the brother as opposed to the haymaker, the church instead of the meadow—indicate the oikotypification of a universal tale according to local rules.

Diachronic variation as seen in the hagiography of Ecgwine and tradition-morphological variation as found in the legends of Juthwara and Sativola are both elaborate bywords for what essentially is a rather simple idea; the production of variants across time and space. These variants are the byproducts of oral transmission; textual vestiges of a performer’s personal artistry interacting with the natural fluctuation of communal credulity. For this reason, identifying variants or adaptations within hagiographic motifs may be considered a useful way of gauging communally-shared folklore.

Reorientating ourselves to appreciate folklore as a participatory activity frees us to envisage other ways in which this primarily oral process may be identified on the written page. We are, by now, familiar with the notion that communities regulate their own lore and filter out unacceptable elements by censoring or sanctioning new variants. [10] But what about acceptable, consensual elements which are so well-known by a community as to be overfamiliar? According to Linda Dégh, commonly-known parts of legends tend to fall out of sight: “[the] common frame of reference absolves the speakers from the need to include minor details of the story or to explain things commonly known within the group.” The disappearance of overfamiliar elements, she argues, “accounts for the brevity and fragmented style of the legend” (Dégh and Vászsonyi 1976, 102). Things that are so well and widely known do not need to be said.

How this applies to hagiography is complicated by the fact these texts are not merely transcripts of oral tradition. Hagiographic legends, even those received aurally, were reshaped by their monastic redactors. The more accomplished hagiographer would naturally have sought to plug the gaps in a fragmented legend so as to render it intelligible to more than just the immediate community. Yet even within the more polished vitae written by professional hagiographers, traces of loose ends and discontinuous elements can be detected. The Vita et miracula S. Kenelmi, for example, gives the appearance of being a coherent, seamless narrative, but a closer reading of the text reveals several slight inconsistencies. Two sisters are initially introduced into the plot, the wicked Cwoenthryth and the virtuous Burgenhild, yet we hear nothing more of the latter. An earlier attempt on the boy’s life is passed over in a single sentence. Cwoenthryth reportedly offered the boy’s tutor “huge bribes and the promise of a share of the [kingdom]” if he were to execute her brother, “[s]ince she could not kill him with poison” (Love 1996, 54). We do not learn how the poison was administered, nor why it was unsuccessful. Finally, despite being an accomplice and would-be profiteer from the crime, the tutor never reappears to receive his comeuppance. Despite the overall coherence of the storyline, several strands are left hanging. These inconsistencies point to a much larger bank of oral legends, familiar to the hagiographer, which he bravely attempted to reconcile into a comprehensive narrative. [11]

Other hagiographers’ attempts were less successful. The first Vita S. Frideswide [BHL 3164] contains a glaring error. Binsey, a wood two miles north-west of Oxford is elided with Bampton, a settlement some twenty miles distant. The early twelfth-century hagiographer describes how Frideswide fled up river to Bampton to evade the clutches of a certain lecherous king of Leicester (Blair 1987, 96-101). There she reportedly took refuge in an abandoned swineherd’s hut in a wood called Binsey, and later cured a young man from Seacourt whose hand had become welded to an axe. The hagiographer clearly cannot have been a local man. Although the medieval settlement of Seacourt did neighbour Binsey, neither was remotely close to Bampton (Biddle 1963). Had he been familiar with the local geography of the Upper Thames Valley he would not have made this mistake. It is likely that the hagiographer was new to the Oxford area, possibly one of the first members of the new Augustinian community (Blair 1987, 83). The vita is stylistically poor which suggests he was not a professional hagiographer. He was most likely drafted in to cobble together a vita for the nascent community’s patrona. He attempted to do this by weaving together fragments from two separate legends, but lacked the skill and geographical knowledge to carry it off. Loose ends and narrative non sequiturs are more detectable in vitae of poorer quality, where the hagiographer made little effort to recast the tales he received or to smooth away their inconsistencies. The fragmented tales would doubtlessly have been familiar to their immediate audience, their arcane nature no barrier to their comprehension. Regrettably the meanings they once encoded for their community are now obscured and they can testify only to places where meaning once existed.

The Vita S. Frideswide demonstrates the importance of transmitting tales through a series of suitable conduits. Once a folktale has passed from the domain of professional or skilled storytellers deterioration begin to set in. This is because it has fallen into the hands of the “wrong type” of tradition bearer or conduit. The multiconduit theory developed by Linda Dégh and Andrew Vászsonyi argues that folklore texts are communicated through society via specific chains of like-minded individuals (Dégh and Vászsonyi 1976). The similarity in their reactions to similar messages qualify these individuals as natural senders and receivers within the conduit. If a folklore text finds its proper conduit, it is transmitted unchanged. But if en route it reaches an inappropriate individual, it will either terminate or survive by undergoing modifications (Dégh 1997, 142). Tradition bearers often have an affinity for a particular genre of folklore. There are different generic conduits: joke conduits which attract comedians, riddle conduits to which those who enjoy posing or solving riddles are drawn, legend conduits for legend enthusiasts and tale conduits for storytellers. An individual with an affinity for telling jokes may be a dreadful teller of stories. People who qualify neither as receivers or transmitters of a tradition are known as “passive bearers” and it is transmission by passive bearers that results in deterioration. Not being the natural choice, these individuals are unskilled, they lack practice and are liable to forget and to make errors under the pressure of performance (Dégh 1995, 207). Deterioration also enters a tradition when the audience is unable to assist a passive bearer and allow errors to proceed unchecked. This occurs once a tale has lost its meaning and relevance to a community (ibid., 199). Having lost interest in enforcing the correct rendition, the audience allow the tale to stray far from its original course, becoming increasingly degenerated.

A combination of these factors must account for the transference of an episode from the legend of Anglo-Saxon-saint Mildrith to that of Augustine the Apostle. In the Vita S. Mildrethe [BHL 5960], a text written by Goscelin of Saint-Bertin in the early 1090s, we find a curious tale about a healing cult on the island of Thanet (Kent) which had sprung up around a footprint Mildrith had reportedly left on a rock when she disembarked from a ship (Rollason 1982, 132-33). Judging from the vita the cult site was popular with the Thanet islanders who used to visit the rock on Rogation Days and were accustomed to curing fevers by mixing the dust they had scraped from the stone with water. The cult site also appears to have been popular with the Benedictine community of St. Augustine’s Abbey, Canterbury, to where Mildrith’s relics had been translated in 1030 (Sharpe 1991). The rock, together with its healing properties, is mentioned in two twelfth-century texts; a set of lessons to be read during the Office on Mildrith’s feast day and the Historia de S. Mildretha, a collection of antiphons and responsories set to musical notation (London, BL, Harley 3908, fols. 38^r and 46^r). The early-twelfth century monks at St. Augustine’s would thus have been familiar with this tradition about Mildrith’s footprint. It was mentioned at least twice in the liturgy used on her feast day and may also have featured in the readings chosen for recital in the refectory as the vita is marked up for breathings (ibid., fols. 1^r-35^r). By the end of the fourteenth century, however, the story had changed dramatically. Although the existence and efficacy of the footprint remain uncontested, the footprint no longer belonged to Mildrith, but to Augustine and the rock no longer lay at Ebbsfleet on the Isle of Thanet, but at Richborough. In his Chronica… de Rebus Gestis Abbatum Sancti Augustini Cantuariae, William Thorne describes the arrival of St. Augustine: “They landed in the island of Taneth in the place called Retesbourgh, where Father Augustine disembarking from the ship stood by chance on a certain stone and left his footmark thereon as if upon mud” (Davis 1934, 4). Thorne was most likely a member of the Thorne family who lived on Thanet, just a mile away from Ebbsfleet, the putative site of Mildrith’s rock. It is possible that he was remembering a story heard during his childhood, a story which had already undergone modification through the mouths of the Thanet islanders. Alternatively, the liturgy used at St. Augustine’s during the twelfth century may have fallen out of use during the intervening centuries, the tale of Mildrith’s footprint no longer rehearsed and its meaning lost for the community. Only vestigial whispers may have remained which over time coalesced around the figure of Augustine. Evidently this tale did not find the right conduit. Devolution set in which, most significantly, went uncontested by the monastic audience. According to the rules of conduit theory, successful transmission demands a narratorial community of congenial persons who listen with pleasure, memorise and perpetuate the tales, and this clearly was no longer the case in Canterbury by the fourteenth century.

Conclusion

The ideas explored in this paper owe a great deal to recent theoretical developments in the field of folklore. Traditionally historians have regarded experiments using theoretical and methodological approaches borrowed from neighbouring disciplines with some trepidation. Caution is rightly counselled against straight importation and such an attempt would be futile if not impossible in this case: certainly, the ethnographical research methods used to analyse contemporary folklore cannot easily be applied to medieval history. Nevertheless, although they may not be good to follow, these conceptual frameworks may be good to think with.

Applying a “performer-centred” approach to hagiographical narratives has proved highly beneficial. It engenders an awareness that variation naturally results from a performer’s personal artistry interacting with communal credulity and is therefore not only the by-product of, but evidence for oral transmission. Thus, I would argue folklore is not characterised by morphological stability and historians hoping to identify folklore (and hence prove orality) in hagiographical texts will not find it rifling through motif indices for matches.

Recognising that textual variation holds the answer for properly historicised readings of these texts opens up a new avenue for research and allows historians to come at these texts anew. Moreover, by understanding some of the underlying reasons behind the production of narrative variation, we can appreciate more fully why certain texts take the form they did. Knowing something about legend conduits, audience participation and performance aesthetics helps us to understand why some tales bear such similarities to others, why others appear so fragmented and incoherent, or why there is such variation in a legend’s subsequent retelling. Reading the texts in this light puts a new complexion on things. It offers insights into monastic storytelling and the way in which institutional traditions were either cherished and revived as was the case at Evesham or else fell from memory as seems to have happened at St Augustine’s. Furthermore, this approach confirms recent assertions that the telling of saints’ legends was a participatory activity shared by both the clergy and the laity. Not only did the clergy appear to have been receptive to stories which originated elsewhere, but they seem to have actively encouraged that contact. It also allows us to go beyond the generality of these conclusions to pinpoint precisely those points of accord, those areas of overlap which united the laity and the clergy. Narratives are never just stories but cultural products whose meaning and significance is negotiated and created at a particular moment between the narrator and his audience. Even the most basic tale encodes information about the world view of these participants. Identifying the kinds of topics where the world views of the laity and clergy coincided, whether they be perceptions of the supernatural, knowledge of healing practices or even concerns about the afterlife, tell us much about the dynamic forces that shaped medieval society.

Abbreviations

BHL

Bibliotheca hagiographica Latina— see “References Cited.”

BL

British Library