Medieval short narratives between languages and cultures: the Miracles of the Virgin Mary


General Introduction

The Miracles of the Virgin Mary as a genre is a multi-lingual literary form comprised of a series of interlinked collections of short, marvellous narratives, which has spread through copying and translation, throughout the Mediterranean world and beyond from the twelfth century to the twentieth, in a bewildering array of languages. The collections, generally involving a semi-canonical core of traditional miracle-tales expanded with locally-based narrations, occupy a borderland between popular culture, learned ideology and political appropriation; form the backbone of a set of ritual and devotional practices; and – in some cases – offer a particularly rich pictorial tradition. The overall tradition offers a rich and vast body of literature, which, in its totality, has not been studied, and whose intertextuality offers a number of interesting problems and resources for further study. The widespread reproduction and re-interpretation of the same miracle-tales further offers an opportunity of studying narratives within Christianity when expressed as a dominant ideology – as with the western European versions – or when invested with dhimmi status – as in the Arabic recensions – or when negotiating and translating material from a different form of Christianity – as with the Orthodox Greek collections.

The project will seek to analyse this complex set of interrelated traditions from three successive standpoints. The first will consider manuscript transmission and the physical distribution of miracle-tales; the second will compare collections and versions, in order to understand the cultural pressures that led to variation and re-elaboration of a set number of miracle-tales; and the third will look at the resulting texts from a narratological point of view, and aim to establish the limits and development of a story within a primarily manuscript culture.

(San Lorenzo del Escorial, ms. T.I.1​, fol. 14r (Cantiga de Santa Maria, no. 7) )

Western Europe offers the most linguistically variegated set of traditions, dominated, however, by the earlier Latin collections. In common with much of the early medieval world, collections of the Virgin’s miracles centred upon a particular shrine were composed, circulated and translated into the vernacular. The undoubted popularity of these collections was rapidly eclipsed in the twelfth century, however, by ‘universal’ collections of miracle-tales, which quite consciously drew stories from different periods and places and sought to de-centre any cult. With a central core of thirty miracles, this new vehicle for marvellous narrations saw multiple collections produced, with expanded the collection to take in new material, re-wrote miracle tales which had become canonical, and re-ordering the miracle-tales into thematic units.

The first stage of the project will be a text-critical study of a selected number of miracles from the core collection, tracing their development across manuscripts, enabling sub-families and recensions to be established, and allowing the evolution of the collections to be precisely identified. Current assumptions about the genesis and development of the collections depend very much upon comparison of contents-lists, but no in-depth text critical study has ever been attempted. Through this tecnique of ‘textual soundings’, the genetic relationships of the individual miracle-tales may be traced, following which a hierarchy of the collections and manuscripts may be established. The stemmatic method will thus not only provide an indication of the ultimate archetype, enabling an edition of all of the core miracle-tales to be produced; but also will isolate and identify individual branches and recensions of the texts and the collections containing them. Much useful work in the intial recording of content-lists has been undertaken for Latin and vernacular collections by the Cantigas de Santa Maria Database (; however, a significant number of manuscripts will require individual consultation as they are insufficiently well-catalogued to identify their contents.

After the central Latin tradition is established, work may begin on the diffusion of the tales beyond the canonical collection. On the one hand, the miracle-tales were included in florilegia of material suitable for sermons (for example, by the late thirteenth-century Bartholomew of Trent, Stephen of Borbon, and Thomas of Cantimpré; or the mid-fifteenth Iohannes Heroult); these texts will depend on certain branches of the main tradition. On the other, through the numerous vernacular translations that were made throughout Europe over from the late twelfth to the early sixteenth century. Dependence of these translations on the Latin textual transmission (or possibly within the vernacular) will be identified.

Miracles of Mary BnP 14th c

Miracles of Mary, BnF, Ms 24541, c. 1330

With a secure basis in textual study, it will be possible to provide a comprehensive understanding of the literary responses to the original text in each case; to evaluate the theological pre-occupations of copyists and translators in effecting changes and altering emphases; and offer a diachronic apprection of the development of the cult of the Virgin throughout Europe. Since many of the translations stand at the beginning of national vernacular literatures, these have been studied thoroughly but usually in isolation, and – as with the Latin collections which have attracted attention only in so far as they relate to England – with an outlook that centres on purely regional associations. It will be possible to draw on this work, much of it detailed and sensitive, whilst re-invigorating it by placing it within a pan-European context, and including works in the discussion which have received merely passing or little attention. At this point, too, Eastern influences on the genesis of the ‘universal’ collection and some of the material will be taken into account.

We shall not, however, simply limit ourselves to the consideration of abstract texts, rendered ideal and dematerialized by critical edition. The constant attention to the manuscript context – the origins of this individual manuscript, the principles behind the creation of this particular exemplar of the tradition – witnessed in the earlier phases of the project can here be brought to bear to ensure that the discussion is situated with the materiality of medieval culture. As a corrollary to this, the depiction of these narratives in manuscript illuminations, stained glass and wall-paintings will also be studied to place the physical elements of Marian devotion into a textual context, and so describe the cult of the Virgin as it developed in western Europe throughout the middle ages, and as it expanded into other territories and other languages, in particular the Arabic and Greek texts that are based upon these traditions, and so the initial comparative analysis may be brought to further fruition through their development across linguistic areas which are rarely brought together, lines of transmission established, and the intricacies of transformations explored.

(Stockholm Royal Library, Isl.Holm.Perg. 4:0 nr. 1, fol. 172r.)

Maríu saga – the life of the Virgin Mary in Old Icelandic, from her immaculate conception to her assumption into Heaven – was probably composed in the early part of the 13th century. It has been attributed to one Kygri-Björn Hjaltason (d. 1237/38), about whom little is known, other than that he was elected as bishop of Hólar in 1236 but died the following year while returning from a journey abroad. He is said (in Guðmundar saga góða) to have written a vita of the Virgin, though whether this was the one we now have is unknown.

Because of the wide range of sources used by its compiler – chiefly the apocryphal gospels and Josephus’ Antiquitates Judicae, Books 16 and 17 – and the way various theological commentaries – by e.g. Augustine, Jerome, Gregory and John Chrysostom, all of whom are named in the text – have been interspersed with biography, the work is unique within the continental medieval Marian tradition.

The saga is found in two recensions, preserved altogether in some twenty manuscripts, the earliest from the first quarter of the 14th century. Most of these also contain miracles, known in Icelandic as jarteiknir or jartegnir (a word originally meaning “token” or “sign”), while collections of Marian miracles, without the saga, are found in a further 25 manuscripts. The only complete edition, that of C. R. Unger from 1871, prints both recensions of the saga and all of the miracles, many in multiple versions.

Altogether there survive around 400 Marian miracle stories in Old Icelandic. Some of these are not found in any of the major continental miracle collections of the Middle Ages and are thus only preserved in Old Icelandic. Nor do any of the existing Latin collections appear to have been used as a direct source for any of the Icelandic ones.

Other Marian works in Old Icelandic include the skaldic poems Líknarbraut (The path to grace), composed around the end of the 13th century, and the Drápa af Maríugrát (The lament of Mary), Vitnisvísur af Maríu (Vision-verses of Mary) and Maríuvísur I–III, all from the 14th century. In the 16th-century manuscript Reykjahólarbók there is a life a Mary based on Low German models.

The first Hungarian poem, the Old Hungarian Lamentations of Mary is dated from the end of the 13th century. The only known variant of this vernacular text is inserted into a Latin codex and is a valued written record of Hungarian literature. No medieval Marian collection has been preserved from the Middle Ages, but there are a number of separate short stories written in the brightest period of the Hungarian codices, which flourished at the turn of the 15th and 16th century. Apart from the list of Floriano Holik—that contains the incipits of 56 Marian miracles (Index miraculorum Marianorum: Indici A. Ponceleti in Anal. Boll. T. XXI. vulgato superaddendus, Budapest 1920)—this Marian corpus hasn’t been thoroughly analysed.

At the Eötvös Loránd University a research group has been founded recently, to build a database which helps to examine the Hungarian exempla of this period ( The project aims to publish not only the vernacular texts but their Latin sources, too.  In most cases the stories are translations of Latin originals (e.g. Legenda aurea, Sermones discipuli…) but usually are transmitted through the Latin sermons of Pelbartus de Themeswar, a Franciscan author from Hungary whose works were printed at the end of the 15th century. The differences found in the Hungarian texts show that oral tradition—mostly throughout the preaching practice—has made a significant influence on the translators. An interesting question to be cleared is the absence of Osualdus de Lasko’s use: he was a published Hungarian author and a contemporary of Pelbartus de Themeswar, but his works don’t seem to have been used in the translations. In addition, the sources of several stories still have to be identified.


”The Private Judgement of György Becsei”: a detail of the wall painting at Zseliz from the 14th century (Photo: Gusztáv Cséfalvay). The scene is related to a Marian short story preserved in the codex Bod—a manuscript written around 1520 probably for Dominican nuns (ELTE University Library, Cod. Hung. 7).

In Greek language, Marian miracle stories are well attested either as individual stories (transmitted as separate entities in menologia, synaxaria and the like) or as miracles embedded in saints’ lives or other hagiographical texts. For the Byzantine period, however, gatherings of her miracle tales were not common. There seems to be only one sizeable collection, namely that of Mary’s miracles worked in the monastery of the Holy Source (Pēgē) on the outskirts of Constantinople (anonymous, 42 tales, later extended and rewritten). A smaller collection originates from the Choziba monastery in Palestine (seven tales).

Miracles at St Mary of the Spring, Painting on wood, Istanbul, Hagia Sophia Museum, 1830

Miracles at St Mary of the Spring, Painting on wood, Istanbul, Hagia Sophia Museum, 1830

Apart from these locally based collections, the Byzantine material consists mainly of individual stories. It is difficult to estimate their number, but we have at least thirty tales of this kind (in several different versions and redactions). Most of these tales are centered round specific Marian icons or relics, viz. the ”robe” and the ”belt” (kept in Constantinople at the Blachernai and Chalkoprateia churches respectively). Throughout the Byzantine period, Mary played a particularly important role in protecting ”her” city from besieging enemies. All of this lore is reflected in the Marian miracle corpus. However, miracle tales seem not to have been gathered into a universalist, canonical collection until after the end of the Byzantine empire. Two such collections exist: one composed by the Cretan Iōannēs Morezēnos in 1599 (”King Solomon’s Bed”, Klinē Solomōntos, 60 tales) and the other by the well-known Agapios Landos (also from Crete), who collected 68 stories to form part of the florilegium ”Salvation of sinners” (Amartōlōn sōtēria), printed in Venice in 1641. Landos probably excerpted Morezēnos’ collection. Interestingly, both Morezēnos and Landos took a number of miracles from a sixteenth-century Italian gathering of miracle-tales, Le miracoli di nostra Donna (Florence 1576), compiled by Silvano Razzi, a Camaldolese monk.

Morezēnos’ collection was not printed until 2007 (and remained therefore relatively unknown), but Landos’ miracle tales were translated into Arabic by the eventual patriarch of Antioch, Macarius ibn al-Za’īm in the second half of the seventeenth century. Despite being an Orthodox translation, manuscripts of this work are present in Melkite libraries, and form the basis of an early twentieth-century continuation (Kitāb ‘Ajā’ib Lūrd) at the Maronite monastery of Sant’ Antonio Abate in Rome.

In the Syriac tradition, two collections of miracle-tales may be distinguished. The first, which was originally composed in Arabic within the Melkite milieu, presented an account of the wonders worked by the miraculous icon at the Saydnāyā monastery, near Damascus.

It circulated in two forms: via a translation into Syriac; and, in the form of a homily attributed to Cyril of Jerusalem, in Copto-Arabic. This latter form was translated into Ethiopian, and, thanks to the interest of the Knights Templars in the shrine, a Latin version was made during the twelfth century, to be followed by a versified version in French in the thirteenth.

The second collection of miracle-tales appear in one recension of a Syriac work, The History of the Blessed Virgin Mary in Six Books, an amalgam of various apocryphal texts current in Syria and Palestine. The miracles are included in a section detailing Mary’s final days upon earth, but their number and order differ from manuscript to manuscript. They fall roughly into two groups: healing miracles and rescue miracles. This collection does not seem to have been known in the West, and its influence on other Eastern traditions, such as Coptic and Ethiopic, is at present unknown.


Miraculous Icon of Saydnaia

In Arabic over one hundred manuscripts of various collections of the Miracles of Mary have so far been identified, ranging in date from the 12th to the 20th cent. A key element in the frequency of the copying of manuscripts was the adoption, at the end of the 14th cent., of the Miracles of the Virgin as liturgical reading for Sundays, Mondays and feasts of the Virgin within the Coptic church.


The Palm Miracle, Story of the Prophets (Qisa al-Abiya) Dublin, Chester Beatty, f. 225

We may divide the various collections as related to their source material: on the one hand, Latin (or western European vernaculars); on the other Greek. One of the earliest forms of the collection was composed between years 1237–89. The translation underwent domesticization: exotic European place-names were made more familiar and there is also some evidence of early cross-over terms developing through vernacularization. Due to the reproduction of some place-names, a northern French original was suggested for at least some of the tales. However, the extempore translation of a Latin text into the vernacular and then translation and transcription into Arabic, would provide another mode of transmission. There is evidence which points to a North African origin for some elements, too. The number of miracles and their arrangement vary, as selections and additions of new material occur during transmission. A means of dividing up manuscript witnesses is to group according to number of miracle-tales: 74 miracles present in 6 mss; the collections of 50 to 68 miracles -7 mss; 32-33 – 4 mss; 14-5 – 2 mss; 9 mss offer totals of miracle-tales unique to themselves from 1 to 43.

Translation from Greek also provided collections with significant numbers of manuscripts. A single representative of the ‘Rhodian’ translation, 42 miracles, is in Paris, BnF, from late 15th cent. The transmission of the 1668 translation copied in the 17th and the 18th cent. is indicated by the mss in Rome, BAV, Paris, BnF, Cambridge, UL Beyrut, Bibliothèque Orientale, Jerusalem, St Anna. Nevertheless, nearly half of the known manuscripts cannot, on present descriptions, be placed even proximately into families, including the earliest manuscript, the potentially 12th cent. BVA, ms. Sbath. 25. There are some 40 other manuscripts which have to be taken in consideration: Rom, BAV; London, BL; Birmingham, Selly Oak; Cambridge, UL; Beyrut, Bibl. orientale; Diarbakyr (Turkey) and in Egypt: Dayr Nasbaih, Dayr Anba Maqar,  St Antony and St Paul at Red Sea, Sinai, St Catherine and Wadi al-Natrun.

The current limited state of our knowledge of the collections and their interrelations makes essential an initial description of miracle-tales (contents, incipit, explicit), and their order in each manuscript. Particular attention should of course be paid to potential cross-overs between the ‘Latin’ and ‘Greek’ collections, together with key indigenous miracle-tales, such as those that occurred at Aufimia and ’Atrib.

At the end of the 14th century Ethiopia, on the far peripheries of Christian Orient, became a terminus station for transmission of the Miracles text, which at that point was an amazing amalgam of literary traditions. Translated from the Copto-Arabic version it was soon enriched with indigenous tales, the process which progressed to recent times, as indicate the current lists which noted over 700 stories. The pioneristic work of E. Cerulli, Il Libro Etiopico dei Miracoli di Maria e le sue fonti nelle Letterature del Medio Evo Latino, Roma 1943 gives a solid ground for further exploration of this development.

The Ethiopian material is unique also in another sense – an important amount of the Miracles manuscripts is illustrated which on the research level involves the problem of a relationship between the text on the one hand and the picture on the other. The origin of this artistic practice needs to be investigated alongside with the methods used by the Ethiopian painters producing a system of extensive narrative picture cycles, displaying in some books up to several hundred episodes and rendering a literary content with great exactitude.


(Enda Eččege, Tämben, Ethiopia, fols 33v-34r; 17th c)