Medieval and Post-Medieval Traditions of Salome’s Icy Death ^†

Abstract

Although the figure of Salome has received much attention from artists and critics alike in the last century and a half, the sources of the legend of her death by ice remain little known. This article traces the origin, development, and transmission of the legend from late antiquity through the early modern period, highlighting its rich presence in medieval Eastern and near total neglect in medieval Western Christian traditions. Drawing on Syriac, Greek, Coptic, Arabic, and Slavic sources, it demonstrates that the legend emerged before the sixth century in the Christian East, where it circulated widely in apocryphal, hagiographical, and historiographical texts. Early attestations, such as the Epistle of Herod to Pilate, already contain core narrative motifs, which later Byzantine authors elaborated into a sophisticated typological narrative of divine retributive justice. In Greek and Slavic traditions especially, the legend achieves rhetorical refinement through the image of decapitation “not by iron but by ice,” while Syriac, Coptic, and Arabic retellings amplify the punitive dimension through the accumulation of often extravagant forms of divine punishment. In contrast, the medieval Latin West preserves only faint and fragmentary echoes of the legend, typically reducing it to brief notices of drowning and largely stripping it of its typological force. Instead, an alternative motif—Salome swallowed by the earth—gains prominence in Western textual and visual culture, possibly facilitated by the conflation of Salome with her mother. The article argues that the fully developed Eastern narrative entered Western Europe only in the sixteenth century through printed Latin translations of Byzantine historiography, particularly the works of Symeon Metaphrastes and Nicephorus Callistus Xanthopoulos. This print-mediated transmission enabled the legend’s broader reception in early modern Europe and its eventual transformation into a folkloric and literary motif, culminating in the nineteenth-century Decadent literature and art.

Oscar Wilde famously ends his play Salome with the princess being crushed beneath the shields of Herod’s soldiers, a decadent and violent tableau precipitated by her perverse kiss upon the severed head of the prophet. However, Georgette Leblanc recalled that in a conversation with Maurice Maeterlinck, Wilde offered a strikingly different version of Salome’s end, one in which Salome, far from remaining a figure of unmitigated depravity, devotes her saintly and ascetic life to bearing witness to John the Baptist and to Christ and merits a death that mirrors that of the prophet whose execution she once demanded: while crossing a frozen lake near the Rhône, she falls through the ice, which snaps shut and severs her head. The head is later seen, crowned with a golden nimbus of a saint, resting upon the silver surface of ice like upon a charger. Wilde, ever the brilliant classicist, remarked that this legend of Salome’s beheading by ice can be found in the writings of Nicephorus, the patriarch of Byzantium (Nicephorus Callistus Xanthopoulus, ca. 1256–1335) (de Saix 1945).

The story of Salome’s death on the frozen lake Wilde recounted—with a saintly twist—is indeed ancient. Its variants circulated widely throughout Eastern Christiandom from late antiquity onward. Locating their origin, however, and precisely tracing their evolution present multiple challenges. Firstly, most of the surviving manuscripts preserving the legend are relatively late and, therefore, far removed—temporally and textually—from the period of late antiquity, when the legend was first taking shape. Secondly, some of the works that preserve it are pseudepigraphic or apocryphal, subject to continual revision and reworking over centuries of transmission; even where they may preserve strands of ancient narratives, those strands cannot be dated or localized with certainty. And thirdly, scribes and later authors could draw on the already circulating written and oral versions of the legend, selectively appropriating, recombining, and repurposing them.

A further complication is the frequent convergence in apocryphal texts of Herodias, the wife of Herod Antipas, and her daughter from her previous marriage to Philip into a single personage. The daughter, who danced before Herod and requested the head of John the Baptist (Matthew 14, 6–11; Mark 6, 21–28), is not named in the gospel accounts; instead, she is referred to as “filia Herodiadis” or “puella.” She was first identified as Salome by the first-century Roman-Jewish historian Josephus in his Jewish Antiquities (XVIII.5.4).1 However, Origen of Alexandria (ca. 184–253 CE), in his commentary on the gospel of Matthew (X.22), clearly asserts that the dancing girl was named Herodias. He writes:

Now the wife of the king of Trachonitis [i.e., Philip], which means a wicked opinion and a wretched doctrine, had borne a daughter who had her mother’s name. Her movements, which appeared graceful, pleased Herod… Her graceful movements are the reason there is no longer a prophetic source among the people… But the dancing of Herodias was contrary to sacred dancing.

(Heine 2018, pp. 56–57)2

Either Origen used a source no longer extant, or else he was rationalizing what was a philological problem. The association of the dancing girl with the name Herodias may have arisen from the variation in pronouns αὐτοῦ and αὐτῆϛ in the early manuscripts of Mark 6, 22, which allows for two different interpretations of the verse. Some of the oldest codices read καὶ εἰσελθούσης τῆς θυγατρὸς αὐτοῦ Ἡρῳδιάδος, which could be interpreted as “and having come in his daughter Herodias…”, while others read καὶ εἰσελθούσης τῆς θυγατρὸς αὐτῆς τῆς Ἡρῳδιάδος, which could be interpreted as “and having come in the daughter of Herodias (herself)…”.3 According to the former formulation, the dancing girl is the daughter of Herod and is named Herodias; according to the latter, she is the daughter of Herodias, but her name is not specified. The presence of the pronoun αὐτοῦ in the textual tradition of Mark 6, 22, may thus explain not only the use of the name Herodias for the dancing girl but also her presentation as a daughter of Herod rather than Herodias.

Given the complexities of the legend’s transmission, this study does not seek to determine the exact origin of Salome’s icy death. Instead, it traces the geographical diffusion of the legend and the evolution of its narrative motifs through a historical–philological approach grounded in comparative analysis of sources across multiple linguistic and cultural traditions. The reconstruction offered here rests on varying degrees of evidentiary certainty, ranging from directly attested textual parallels to more tentative inferences based on recurring narrative configurations. Drawing on Syriac, Greek, Coptic, Arabic, Slavic, and Latin texts, the study examines issues of manuscript transmission, dating, and intertextual relationships. In assessing relationships among these texts, the analysis distinguishes, where the evidence permits, between direct textual dependence supported by close verbal correspondences, mediated transmission, and broader motif–level or typological parallelism. Where such distinctions cannot be securely established, the discussion deliberately adopts more cautious formulations to reflect varying degrees of evidentiary certainty and to avoid overstating claims of influence.

Based on its findings, the study advances the hypothesis that, although rhetorically and typologically elaborated versions of the legend circulated widely in the medieval East, it remained largely unknown in the medieval West, where its absence was only partially compensated by vestigial references to Salome’s drowning or to her being swallowed by the earth. The complete version of the legend, recounting her death on the ice, reached Western Europe only in the sixteenth century with the publication of Latin translations of Greek historians. It was therefore not until the early modern period that the legend entered Western cultural consciousness, enabling its subsequent literary elaborations in the late nineteenth and early twentieth centuries.

Previous scholarship has not fully addressed the problem of the legend’s diffusion. While scholars of Eastern Christianity have occasionally drawn attention to it, studies of Salome in the medieval and modern West have tended to focus on the image of the girl dancing before Herod. The legend of her icy death is typically passed over in silence or mentioned only in passing. This study does not attempt a comprehensive review of that scholarship; rather, it deliberately sets aside earlier treatments that do not engage the question of the legend’s transmission in order to focus on reconstructing its dissemination across linguistic, cultural, and chronological boundaries. It thus offers the first systematic attempt to trace its dispersal across the Christian world and to account for its relatively late emergence in Western Europe.

1. Eastern Christianity in Late Antiquity and the Middle Ages

The origins of the legend of Salome’s death on the ice can be situated in Eastern Christianity by the sixth century. The legend is invoked in the Syriac version of the apocryphal Epistle of Herod to Pilate (Ep. Herod Pil.—Syr.) preserved in a manuscript dated to 587, which provides a secure terminus ante quem for both the letter and the legend (Geerard 1992, no. 67; Burke 2024b, no. 785).4 Its survival in Syriac, Greek, Coptic, Arabic, and Slavic texts suggests a fairly wide circulation in the Eastern regions.

In the letter, Herod reveals to Pilate the reasons for his profound grief, and he begins by recounting the tragedy that befell his daughter named Herodias, like her mother. The relevant passage reads as follows:

For as my daughter, who was dear to me, Herodia, was playing on a deep (pond) of water which was frozen over, the ice broke under her, and her whole body went down, and her head was cut off, and remained on the surface of the ice. And lo, her mother is holding her head on her knees in her lap, and my whole house is in great sorrow.

(Wright 1865, p. 12)5

The letter speaks of Herod’s daughter, rather than Herodias’s, and calls the girl Herodias rather than Salome (cf. Origen of Alexandria, quoted above). Despite its brevity, this account comprises a number of narrative motifs that recur also in later retellings: the girl playing on a frozen body of water (here not specified), the ice breaking, her the body going underwater, her head being severed and resting on the ice, her mother holding it on her knees in her lap, and the household witnessing the events in great lamentation. The letter does not explicitly describe the manner of the girl’s decapitation but implies that the head was severed by ice. Its resting place on the icy surface is clearly meant to evoke the image of John the Baptist’s head on a charger and to emphasize the symmetry of retributive justice.

The Greek version of the Epistle of Herod to Pilate (Ep. Herod Pil.—Gk), currently known from a single fifteenth-century manuscript,6 is much later and significantly reshapes the episode:

My beloved daughter Herodia was killed while playing by the water, when it flooded over the bank of the river. For suddenly the water rose up to her neck, and her mother grabbed her by the head to keep her from being swept away by the water. The head of the child was severed so that my wife held only the head, while the water took the rest of her body. And so my wife held her head on her knees, weeping, and all my household fell into incessant grief.

(Ehrman and Pleše 2011, p. 525)

Although it mentions a flooding river—so a body of water—it makes no mention of ice as the instrument of the girl’s decapitation. Instead, it is her mother who, attempting to save her from being swept away, grabs the girl by the head and apparently causes the head to be severed, with the rest of the body taken by the river. However, the images of Herod’s wife holding the head of her daughter in her lap and of the whole household in grief are closely reminiscent of those in the Syriac letter.

A similar account appears also in the Greek List of the Apostles and Disciples ascribed to Dorotheus of Tyre (List Doroth.), who, according to the text itself, was martyred under emperor Julian (361–363) (Burke 2022).7 Although the internal references in List Doroth. might suggest that it antedates the late sixth century, it was probably compiled much later from pre-existing materials. According to Tony Burke, “[t]he full compilation was likely assembled in the eighth century; certainly before 810/811, when Theophanes the Confessor composed his Chronicle, which mentions Dorotheus and his list of bishops” (Burke 2022). The text claims that Dorotheus wrote this document in Latin, and a certain Procopius translated it into Greek (Pseudo-Dorotheus 1907, pp. 143, 159–60). Latin copies of List Doroth. do indeed exist, but the legend is attached to the document only in the Greek version,8 which survives in multiple manuscripts dating from the tenth century onwards. It reports the legend as follows:

[About the daughter of Herodias.] During the consulship of Galba and Sylla, under these consuls, as the Gennesaret Lake was icebound by frost, Herodias’s daughter went out for pleasure onto the frozen surface. But when the frozen surface broke into pieces, her body was swallowed by water, but her head was cut off and remained above the frozen surface. And Herodias took her daughter’s head on her knees and wept, and confessed that she endured this because she had asked for the head of John the Baptist.

(My translation)9

This version, too, carries echoes of the Syriac letter of Herod: the girl, although here she is the daughter of Herodias, plays on the frozen Lake Gennesaret, the ice breaks, the girl’s body sinks. Like the Syriac letter, List Doroth. is not explicit about the manner in which her head was severed but simply states that it was cut off and remained on the surface of the ice. It concludes, again like the letter, with her mother placing it on her knees. The final confession of the girl’s mother that she brought the tragedy on herself by asking for the head of John the Baptist may be a reflex of a similar confession uttered by Herod in his letter. In fact, some relationship between the two texts is also suggested by a later reference in both to Pilate, a gentile, becoming Herod’s heir.10

Subsequent Byzantine historiography preserves abbreviated echoes of versions similar to the ones described above. In the eleventh-century Compendium historiarum, Georgius Cedrenus tersely notes that

they say that one, playing on the frozen surface as the lake ice broke, went down, and the head was wedged in tightly, and the entire body went down into the depths, but the head remained on top of the frozen surface.

(My translation)11

The place is still a lake, but it is not otherwise specified, and the final tableau with Herodias holding her daughter’s head is also suppressed. The legend is here condensed into a series of key motifs.

The details of the legend continued to fluctuate in Greek and Slavic apocrypha about Zachariah and John the Baptist. They include The Martyrdom of Zechariah (Mart. Zech.), whose earliest Greek text is preserved in an early fourteenth-century codex (Veale 2017, no. 509; Geerard 1992, no. 180.1/180.4; Burke and Veale 2023, pp. 147–49), and at least two different Greek versions of The Decapitation of John the Forerunner (Decap. Bapt.): version A first attested in the tenth or eleventh century and version B about a century earlier (Burke 2017, no. 400; Geerard 1992, no. 180.2; Burke 2023, pp. 158–75). Although the manuscripts provide useful terminus ante quem for the composition of these apocrypha, they shed little light on their relative chronology. Slavic copies of these apocrypha were typically derived from the Greek and belong to the fifteenth century or later, with the exception of one fourteenth-century copy of Mart. Zech. In all these apocrypha, one Euripios, an otherwise unknown disciple of John, claims to be the author.

In Mart. Zech., the legend is reported as follows:

It happened on a winter day, the daughter of Herod was playing near a frozen well, and slipping, she fell into it. And her trunk was caught and her entirety could not be seen because the ice caught her, but only her head stood out. Those nearby and her household ran to tear her free by force. Her head was torn free but her body went out of sight into the well, and many who felt around did not find (it). Then, taking up her head, they carried (it) to Herod and Herod placed (it) on his lap with much sorrow.12

In this iteration of the legend, Herod’s (rather than Herodias’s) daughter falls into a frozen well rather than a natural body of water; only her head stays above the ice. As in Ep. Herod Pil.—Gk, she is not decapitated by ice, but her head is torn from her body in an attempt to rescue her, although in contrast to the Greek letter of Herod, it is not her mother who causes her death but “those nearby.” Eventually, the head is placed not in Herodias’s lap but in Herod’s, a detail characteristic of this group of texts.

The younger of the two versions of Decap. Bapt., related to Mart. Zech., gives a similar story, with only minor differences:

In that time, the daughter of Herod, on a winter’s day when it was cold, was dancing near a frozen well of water, (and) falling, she <sank> into the water. Those nearby, wishing to pull her out, cut off her head and her trunk went down.

When Herod sat down, the head of Herodias was brought, and he put it on his lap. He began to cry and say, “O righteous water, of unrighteous death most sharp! …”13

In this version, the girl is dancing rather than playing, and the bystanders cut off her head rather than tear it off. After receiving his daughter’s head in his lap, Herod utters a lament, which is not integral to the legend in Mart. Zech., although sentences from it (e.g., an apostrophe to the righteous water) appear in it later.

Decap. Bapt. B, the older of the two recensions of Decap. Bapt., is much shorter and contains only the gist of the events:

She was beheaded by the frozen water. And her head was brought to her father but her body remained below. O sharp water of iron, decapitating a head as a condemnation of unbelievers…

(Burke 2023, p. 174)

This abridgement stands in contrast to both Mart. Zech. and Decap. Bapt. A, as it preserves the original version of the girl’s death, decapitation by ice. The mention of Herod receiving her head, however, brings it back in line with version A. The same condensed summary is attested also in two printed Slavic versions of Decap. Bapt. B.14

From the late tenth century onwards, the legend of Salome’s death on the ice was rapidly spreading in the Greek-speaking world through the work of Symeon Metaphrastes (d. ca. 1000), a hagiographer commissioned by emperor Basil II (976–1025) to compile an authoritative collection of saints’ lives for official liturgical use. Symeon relates the story in the Hypomnema on John the Baptist (Hypom. Bapt.), a part of his Menologion. The textual tradition of this work is well attested: its oldest manuscript dates to the tenth century, and more than a dozen survive from the eleventh.15

Symeon typically worked by revising and amplifying earlier materials. Although his version of the legend shares with Ep. Herod Pil. the combination of icy submersion and the separation of the head from the body, no direct textual dependence can be demonstrated. However, the correspondences suggest that both accounts belong to a common narrative tradition. His version reads:

However, the daughter (who has been said to have perished before her mother), while going from one place to another and having to cross a river in wintertime, since it was ice-bound and frozen, she was walking across it on foot. But the ice having burst all around—not without the aid of God—and she, the wretched one, sank down at once, indeed, down to her head, with convulsions as if even then dancing, not on earth but in water; but the head, shattered by ice and terribly severed, was left above the frozen surface, a marvelous sign of God’s dispensations; it was separated from the rest of the body not by a sword but by ice. But these things happened in this way, and that infamous and abominable head lay there in the sight of all, leading the onlookers to a remembrance of what she had done.

(My translation)16

Metaphrastes changes the narrative situation from playing or dancing on a body of water bound by ice to a more purposeful crossing of a frozen river. To highlight the parallel with the girl’s dance at Herod’s feast, he describes the dance-like movements of her drowning body and ends with an explicit description of the ice severing her head, which then rests on the icy surface, cut off not by a sword,17 as John’s head, but by the shards of ice. This drama played out for the benefit of the onlookers, whom it was meant to remind of the dancer’s wicked acts. Neither Herod nor his wife are mentioned as receiving her head, presumably so as not to distract from the parallel with the decollation of John the Baptist and the message of retributive justice.

About three centuries later, Symeon’s account was itself re-worked by the Byzantine historian Nicephorus Callistus Xanthopoulus (ca. 1256–ca. 1335). The account in his Historia ecclesiastica reads:

But as for his daughter’s death—it is worth recounting—it was like this. Having to travel to a certain place in this way in wintertime and to cross a river, since it was ice-bound and frozen, she was going over the surface on foot. But the ice having burst all around (surely, this did not happen without the aid of God), she sank down at once, indeed, down to her head; and, swelling, she began to dance, bending fluidly not on land but in the water; but her head was frozen solid by the cold; then, it was shattered, and it was separated from the rest of the body, not by a sword but by ice. Upon the frozen surface, it was dancing its dance of death. And the abominable head lay in the sight of all, leading the onlookers to a remembrance of what she had done.

(My translation)18

Nicephorus borrows the legend from Metaphrastes: he retains the references to the crossing of the frozen river on foot, God’s approval, dancing-like movements in water, the ice—rather than the sword—cutting off the head, and the explicit comment on the spectacle reminding the onlookers of the girl’s earlier actions. However, he replaces the image of the head lying on the ice with the one of the head dancing (bounding and bouncing?) on the ice; that is, in Nicephorus’s account, both the body and the head re-enact the deadly dance performed before Herod.

The legend of Salome’s decapitation by ice was also retold by medieval Syriac, Coptic, and Arab writers, inspired by Ep. Herod Pil.—Syr. or some similar traditions. Those retellings often embellish the story with additional motifs, all the while enhancing the theme of divine justice, retribution, and vengeance. One of the earliest such retellings is preserved the ninth-century Commentary on St. Matthew by Isho’dad of Merv, bishop of Ḥadatha (near modern Mosul, Iraq), who flourished ca. 850:

Now the damsel, after she had taken the head of John in a charger, and brought it to her mother, returned to the guests, that with new kinds of her dancing she might pay the wages of her request. There was a lake at the side of which the dining hall was fixed; and she went upon the ice in order to dance and to shew the excellence of her performance, and amaze the beholders, when suddenly that place was opened from below her, and she was swallowed up as far as her neck, and a great fish was commissioned from God for the revenge of his death; it swallowed up her body, and when by every means they strove to rescue her, they could not; and as soon as they cut off her head with the very sword with which John was murdered, the Earth threw it up without any man lifting it; and while the head of John had been put before her mother, and she was striking it on the earth and mocking it, as “Where is thy mouth that embittered our lives?” there was put also before her the head of her daughter; and immediately from much weeping by one angelic operation her two eyes dropped and they fell upon the head of her daughter and [on that] of John.

(Isho’dad of Merv 1911)

Isho’dad of Merv alters the circumstances of the fateful dance, situating it on the frozen lake beside the feast hall immediately after the girl presented John’s head to her mother. Her motivation is new: she dances on ice to show off her excellence and to amaze the onlookers. Isho’dad then adds yet another twist to the story: as the ice breaks from below, she is swallowed up to her head by a great fish exacting divine retribution. The attempted rescue, encountered also in some Greek versions, ends with the rescuers decapitating the girl with the very sword used to decapitate John and with the Earth throwing up her body.19 Her head is brought to her mother but, instead of being placed in her lap, as in Ep. Herod Pil.—Syr., it is placed beside the head of John, which her mother was reviling. From weeping, her mother’s two eyes miraculously fall on the two heads before her.20 Isho’dad’s version of the legend appears to combine elements associated with three narrative patterns: decapitation by ice, swallowing by a creature of the water, and execution with the sword used against John the Baptist. Whether these elements derive from separate written sources or from a fluid oral tradition cannot be determined, but their convergence here produces a markedly intensified, verging on excessive, vision of retribution.21

Isho’dad’s version of Salome’s death may have been the source of the Syriac Book of the Bee, written by Solomon of Akhlat, bishop of Basra, around 1222, although it is not impossible that Solomon used the same source(s) as Isho’dad. His account runs as follows:

Then she went out to dance upon the ice, and it opened under her, and she sank into the water up to her neck; and no one was able to deliver her. And they brought the sword with which John’s head had been cut off, and cut off hers and carried it to Herodias her mother. When she saw her daughter’s head and that of the holy man she became blind….

(Budge 1886, pp. 90–91)

Salomon preserves the core elements of Isho’dad’s narrative, but he also abridges the story by cutting out the references to the girl being swallowed by a fish and to her body being thrown up by the earth. He follows the legend with a brief description of her mother’s physical deterioration and her possession by Satan.

The narrative excess present in Isho’dad’s version is evident also in the Arabic Life of John the Baptist ascribed to a certain Serapion (Life Bapt. Serap.), usually identified with an Egyptian bishop, Serapion Scholasticus, bishop of Thmuis in the second half of the fourth century (Burke 2014, pp. 277–78). However, as the recent translator of Life Bapt. Serap. notes, it is unlikely that the text as we have it today is the actual ancient composition; rather, what has survived is a medieval compilation of earlier materials, put together before the fourteenth century in Christian Egypt, probably in Coptic, and then translated into Arabic. All its manuscripts are written in West Syriac script (Garšūnī), and all are late, the earliest dating to the fifteenth century (Čéplö 2016b, p. 275–76; Čéplö 2016c; Geerard 1992, no. 183). Pseudo-Serapion’s account of the dancing girl’s fate is preceded by that of her mother:

As for Herodias [i.e., mother], her eyes were pulled out from her head and fell on the ground. Her room collapsed on top of her, the ground opened its mouth and its throat swallowed her, and then she sank to the depths of hell, still alive. Herodias’s daughter went mad and broke all the vessels that were there at the feast. In her madness, she went to a frozen lake and danced on it. The Lord ordered the ice under her to break and the lake swallowed her. Soldiers tried to pull her out and could not, because the Lord did not want her to be rescued. Finally, they cut off her head using the sword with which holy John was killed. At that very moment, a whale appeared and threw her out of the lake, dead. May God have no mercy on her! Immediately after that, Herod suffered a stroke in front of his dinner companions.

(Čéplö 2016b, p. 289)

Like Isho’dad, pseudo-Serapion situates the girl’s last dance still during Herod’s feast, but her motives for dancing on ice are different: in Life Bapt. Serap. she dances on ice because she goes mad rather than to gain further praise. The ice breaks at the Lord’s behest, as in the accounts by Symeon Metaphrastes and Nicephorus. There follow the motifs of the failed rescue and the beheading with the sword that beheaded John the Baptist, as in Isho’dad and Solomon of Basra. The whale casting her out of the lake evokes the fish in Isho’dad’s account, as does the fate of Herodias sinking into hell.22

The motif of the girl’s madness, encountered in Life Bapt. Serap., must have been current in Egypt for it appears also in the fourteenth-century Arabic apocryphon On Herod and John the Baptist, translated, like Life Bapt. Serap., probably from Coptic (Čéplö 2017, pp. 295–319; Čéplö 2016a). In this version, however, it is Herod himself who beheads the girl:

As for her daughter, the devil entered her and she broke all the vessels in the house…

As for her daughter, she was mindlessly running around and hallucinating, and her hair fell from her head from all the running around, so the king cut her head off and ordered his companions not to tell anyone of this secret. But Herodias’ shame became known to everyone and the smell of her daughter spread.

(Čéplö 2017, pp. 306–7)

The accounts of Salome’s death described above suggest that two distinct narrative tendencies emerged across the Eastern Christian traditions. In Greek and Slavic texts (such as those associated with Symeon Mataphrastes and Nicephorus Callistus Xanthopoulus), the dominant motif is a highly stylized, typological inversion: the girl crosses or plays upon frozen water, the ice breaks, and her head is severed by the ice itself—explicitly “not by iron but by ice”—so that her death forms a symmetrical counterpoint to John’s beheading and becomes a visible sign of divine retributive justice. By contrast, Syriac, Arabic, and Coptic traditions tend to multiply and intensify the mechanisms of punishment: alongside or instead of icy decapitation, she may be swallowed by water, fish, or whale; go mad; or be beheaded with the very sword that killed John, producing a cumulative and dramatic vision of retribution that is less concerned with elegant symbolic inversion than with overwhelming, almost apocalyptic divine vengeance.

The scope of the narrative expands over time. What begins as a brief explanatory episode embedded in an apocryphal letter becomes, in later centuries, a developed moral drama incorporated into apostolic lists, martyrdom accounts, biblical commentary, liturgical menologia, and universal histories. In Syriac and Arabic retellings especially, the story accumulates additional punitive motifs—fish or whale, the swallowing earth, madness, demonic possession—layered onto the earlier icy death, amplifications that heighten the rhetoric of divine vengeance.

2. Western Christianity in the Middle Ages

While the eastern Mediterranean cultivated and elaborated the legend across a wide range of languages and genres, the Christian West appears to have remained largely unaffected by it.23 So far no datable evidence has been brought forth to suggest that the Eastern narrative of Salome’s death on the ice circulated in Latin Europe before the twelfth century. Even thereafter, attestations are sparse—no more than four or five across the entire medieval period; they are highly abridged, and they never achieve the narrative density or theological emphasis characteristic of their Eastern counterparts.

The most important evidence for the limited presence of the legend in Western Europe from the twelfth century onwards is the Latin version of the Epistle of Herod to Pilate (Ep. Herod Pil.—Lat.), which survives in three closely related manuscripts, two of which are dated to the first half of the twelfth century.24 The legend is summarized there as follows:

Indeed, you should know that my daughter, Herodias, as she was playing on frozen water, fell in. As soon as she fell in, her head was detached from her body, so much so that my wife may hold the very head in her lap. You should be aware that, for this reason, all my household is in mourning.

(My translation)25

This account of Herod’s beloved daughter’s death opens the letter rather abruptly. The narrative is very close the Syriac version, except that the Latin text omits the mention of the head remaining on the surface of the ice. It ends with her head in her mother’s lap (in Syriac, it rests on her mother’s knees in her lap). Both accounts conclude with a reference to Herod’s entire household engulfed in grief.

However, there is no evidence that the thirteenth-century Dominican Jean de Mailly drew directly on Epistula Herodis Pilati—Lat. when recalling the legend in his Abbreviatio in gestis et miraculis sanctorum; his compressed formulation is more consistent with indirect transmission through some oral or written précis of the legendary material. He reduces the episode to a single sentence: “[C]oncerning the daughter of Herodias, it is said that when she was walking on ice, the ice melting under her, and she drowned” (my translation).26 Moreover, since his version emphasizes the drowning but does not mention the decollation, it no longer mirrors the death of John the Baptist. The allusion to the legend has lost its typological effect, which left it less attractive as an illustration of divine retributive justice. A few sentences later, Jean invokes a different account of Salome’s death, when he writes “But it is read in chronicles that the earth swallowed alive Herodias’s daughter who danced for the decollation of Saint John” (my translation),27 effectively undermining the imaginative force of the previous explanation of Salome’s death.

Jean de Mailly’s statement about the girl’s drowning is echoed, albeit with altered wording, by Jacobus de Voragine (ca. 1230–1298) in the chapter “On the Beheading of Saint John the Baptist” of his Legenda aurea: “As for her daughter, she was walking over an icy pond when the ice gave way under her and she was drowned, though one chronicle says that the earth swallowed her alive. This is understandable, since it is said of the Egyptians who were drowned in the Red Sea: ‘The earth swallowed them’“ (Jacobus de Voragine 1993, pp. 138–39).28 In his explanation, Jacobus interprets one account of the girl’s death in terms of another by means of a reference to Exod. 15:12, merging the two different accounts into a single one. There is no evidence that Jacobus was familiar with the motif of the dancing girl’s decapitation by ice.

Together with Ep. Herod Pil.—Lat., these two brief allusions to Salome’s drowning effectively exhaust the Latin references to the legend that saw such wide diffusion in the East. Vernacular references in Europe are equally scarce, and so far only two have come to light. Neither of them goes beyond the short formulations of Jean and Jacobus. The first occurs in the thirteenth-century Middle High German Passional, which reads,

the evil one, the foolish one

once went upon ice;

she did so for amusement.

See how the ice broke underneath!

the maiden fell in and drowned.

(My translation)29

Again, this brief and rudimentary allusion betrays no direct acquaintance with the fuller narrative elaborations of the legend.

An equally compressed reference appears in the fourteenth-century Middle Low German Große Seelentrost: “Wretched, he [Herod] died there with his wife, and the daughter had to walk across the ice; the ice broke, and she downed” (my translation).30 Here again, the exclusive emphasis on drowning points to one of the aforementioned Latin summaries as the source and suggests no awareness of the more elaborate narrative in which decapitation by ice mirrors the beheading of John the Baptist.

This relative marginality of the legend of the dancer’s death on the ice within the legendary landscape of medieval Western Europe is striking. Given the prominence of John the Baptist in the Latin West, one might expect it to have circulated more widely. It did not. That is because the fully developed narrative never entered the main Latin imaginary of the gospel villains, and the sparse Western attestations do not amount to an independent elaboration of the story. With the exception of Ep. Herod Pil.—Lat., they mostly preserve a highly abbreviated and redirected form of the tale—a muted echo of material long established and rhetorically shaped in Byzantine and Eastern Christian contexts.

If the legend is largely absent from Western textual sources, it is likewise missing from the iconographic repertoire of Western artists. While the scene of the girl’s dancing before Herod and receiving a charger with the head of John the Baptist became ubiquitous in medieval Western art, her own decapitation by ice is conspicuously absent. To date, only a single miniature has been interpreted as depicting Salome drowning, namely, in the fourteenth-century Carmelite missal reconstructed by Margaret Rickert. However, this interpretation remains uncertain since Rickert identifies as water what may instead represent the earth engulfing Salome. Whether Rickert’s interpretation is correct or not, the miniature plausibly reflects the influence of Jean or Jacobus, both of whom mentioned two versions of Salome’s end (Rickert 1952, pp. 107–8 and plate XVI(b)).

The motif of the earth swallowing Salome is depicted unambiguously in a miniature illustrating a French historical compilation based on the Compendium historiae in genealogia Christi by Petrus Pictaviensis (ca. 1130—1205). In the manuscript, dated 1300–1310, the miniature shows a woman, her distress evident from her upraised hands, being engulfed by what appear at first glance to be green waves. Yet these “waves” support a green tree, strongly suggesting that they represent not water but the jaws of the earth. The text beside the miniature clarifies the scene, “the girl who danced before king Herod who held Galilee, through whom the Baptist was decollated, was swallowed by the earth” (my translation).31

The alternative version of Salome’s fate encountered in the Latin West, namely the idea that she was swallowed by the earth, may have originated in traditions concerning her mother, since mother and daughter were frequently conflated under the shared name Herodias. It is the mother, rather than the daughter, who is swallowed by the earth in early Coptic tradition and in later Arabic sources.32 One of the earliest references to the dancing girl herself being swallowed by the earth occurs in Chronographia tripertita by Anastasius Bibliothecarius (ca. 810—ca. 878), “the earth swallowed alive the girl who danced for the decollation of John the Baptist.”33 In Greek, it surfaces in Georgius Cedrenus, who invokes it before reporting the girl’s death by ice.34 In the later Middle Ages, such notices become quite frequent in the West. In the thirteenth century, Sicardus Cremonensis states that “the earth swallowed alive the girl who asked for the head of John” (my translation);35 his contemporary, Matthew Paris, echoes this wording, “and the earth swallowed alive the dancing girl” (my translation);36 in the fourteenth century, Ranulph Higden follows suit, “but the earth swallowed the girl who danced” (my translation);37 and the annals of Worcester Priory from the same century record that, “Herod the tetrarch decollated John the Baptist; the living earth swallowed alive the girl who danced for his decollation” (my translation).38

3. Early Modern Europe

The relative marginality of the legend about Salome’s icy decapitation in the medieval Latin West renders what happened after the invention of print all the more significant. From the sixteenth century onward, the accounts of Symeon Metaphrastes and, especially, Nicephorus Callistus Xanthopoulos were translated into Latin and repeatedly printed, thereby entering Western Europe in a stable and widely accessible form. The medieval cursory reflexes of the legend were replaced by its full, authoritative account and made accessible to a transregional readership literate in Latin. This shift was not merely one of scale but also of status: through translation and print, the legend was recontextualized within the recognized corpus of ecclesiastical historiography and hagiography, where it could be received not as marginal but as supplemental to the biblical narrative. The legend began to attract attention. It was this post-medieval, translated, and print-mediated reception of Salome’s death on the ice that enabled and motivated the Salome folklore recorded in the eighteenth and nineteenth centuries in Europe, as well as the marked spike in the reception of the legend at the turn of the nineteenth and twentieth centuries. The ensuing section will, therefore, examine the mechanisms of this printed transmission and the gradual reconfiguration of Salome’s fate in the Western imagination.

The first Greek work containing an account of Salome’s death to be translated into Latin and printed was Compendium historiarum by Georgius Cedrenus, published in a dual language edition (Greek and Latin) by Wilhelm Xylander in 1566 (Cedrenus 1566). Cedrenus’s version was brief but, set in a historiographic context, it conferred on the legend a degree of legitimacy. Hypomnema on John the Baptist by Symeon Metaphrastes, which contains a fully developed version, was issued in Latin first by Aloisius Lipomanus in Venice in 1568 and shortly thereafter by Laurentius Surius in De probatis sanctorum historiis in Cologne (Lipomanus 1558, fol. 201r; Surius 1573, p. 969). By inserting it into widely circulating compilations of sacred history and hagiography, they began to integrate the legend with the established body of religious reading and instruction.

It was, however, Nicephorus’s Εκκλησιαστική Ιστορία, translated into Latin by Johannes Langus as Ecclesiasticæ historiæ libri decem et octo, that proved most influential in popularizing the legend in the West. The Latin Nicephorus was printed in 1553 in Basel (Nicephorus Callistus Xanthopulus 1553).39 Subsequently, it was reissued in Basel in 1561; in Paris in 1560, 1562, 1566, 1573 and 1574; and in Frankfurt in 1588. This list probably does not exhaust the sixteenth-century editions, and the Historia continued to be printed in the seventeenth century as well. Such repeated publication ensured not only wide geographical dissemination but also the stabilization of a particular, rhetorically elaborated version of the narrative. In this form, the story of the dancing girl crossing the frozen river and of her death-dance on the ice and in water as a manifestation of divine retribution became part of the shared repertoire of religious culture in Europe. Its authority derived in large measure from its historiographical framing: transmitted as part of an “ecclesiastical history,” the account could be read as an extension of the biblical past rather than as a merely apocryphal embellishment.

Nicephorus’s version was taken up by numerous early modern authors. Juan Bonifacio (1538–1606), in his Institutio christiani pueri, adolescentiaeque perfugium, quotes and acknowledges Nicephorus as his source for the story of Salome’s decapitation by ice (Bonifacio 1607, p. 27); Jan Buys (1547–1611) reproduces it in Panarion, hoc est arca medica variis diuinae Scripturae priscorumque Patrum antidotis aduersus animi morbos instructa (Buys 1608, p. 26); Cornelius à Lapide, includes it in his commentary on Mark 6, 26;40 and Georgius Fabricius incorporates it into Virorum illustrium seu historiae sacrae libri X recogniti et aucti (Fabricius 1606, p. 264).41 In these contexts, the legend serves as a moral exemplum, readily adaptable to pedagogical, exegetical, and devotional purposes.

Occasionally, authors added local colour to the rapidly spreading legend. Thus, in his commentary on the Chronicle of Pseudo-Dexter—a work composed in fact by Jerónimo Román de la Higuera’s (1538–1611) and printed in 1594—Franciscus Bivarius identifies the frozen river in which Salome perished as “Sicorim, flumen Ilerdæ,” adding that the river is “vulgo nunc Segre dictus” (Dexter 1627, p. 34).42 Such attempts to anchor the narrative in specific geography suggest a desire to enhance its plausibility and immediacy for contemporary readers. Within a few decades, the image of divine justice enacted through the freezing and severing of Salome’s head had become permanently inserted into Europe’s religious heritage.

The process of diffusion did not remain confined to Latin. It continued in the vernacular as Nicephorus’s account was either translated along with the rest of his Ecclesiastical History or incorporated as an instructive episode into pious vernacular writings. For instance, the first French translation of Ecclesiastical History, rendered from the Latin of Johannes Langus, was printed by Sebastien Nivelle in 1567 (Nicephorus Callistus Xanthopulus 1567). More commonly, the episode of Salome’s death was incorporated into larger devotional collections. Thus, it was included in Alban Butler’s The Lives of the Fathers, Martyrs, and Other Principal Saints, first published between 1756 and 1759 Cf. (Butler 1799, p. 543)43; it was quoted in extenso by Abraham a Sancta Clara (Johann Ulrich Megerle, 1644–1709) in his Abrahamische Lauber-Hütt, published posthumously between 1721 and 1723 (Abraham a Santa Clara 1722, pp. 99–100); and it was disseminated especially through Martinus von Cochem’s Großen Leben Christi, which went through more than two hundred editions and became the cornerstone of Catholic piety in German speaking lands in the eighteenth and nineteenth centuries (Kretzenbacher 1969, pp. 193–94). In these contexts, the legend was detached from the historiographical contexts and presented as an edifying illustration of divine retribution.

The speed and breadth of this early modern diffusion suggest that, had the fully developed legend been available during the Middle Ages, it would likely have found far wider expression in hagiographic, theological, and devotional literature. Its earlier marginality must, therefore, be attributed not so much to a lack of narrative appeal as to the contingencies of transmission. Once introduced through the medium of print and authorized by association with respected historiographical sources, the legend proved highly adaptable. Eventually, it even moved beyond the sphere of learned and devotional writing into the world of storytelling and folklore, where it sometimes merged with older traditions concerning the nocturnal wanderings of Herodias,44 and into literary culture, where it inspired the imaginations of poets, novelists, and playwrights, particularly during the Decadent period.45

The legend of Salome’s death on the ice, mentioned by Wilde in his conversation with Maeterlinck, has thus a long history, though not one rooted in the medieval West. The legend was imported from the East in the early modern period, and it was only then that it truly entered the Western cultural horizon—first within leaned and devotional discourse and later in folklore and secular literature. Its reception reached an apogee in the late nineteenth and early twentieth centuries. Wilde’s fascination with it was shared by Charles Buet in L’Expiation de Salomé—legende (1879), Apolinaire in “La Danseuse” (Trois Histoires de Châtiments divins, 1902), and Michael Field in “A Dance of Death” (Poems of Adoration, 1912), among others who transformed the legend into literary works of art. Salome’s sainthood, however, appears to be a flourish of Wilde’s own imagination (Ogane 2011, p. 155; Jordan 2012, p. 10).

4. Conclusions

In light of the evidence presented above, the legend of Salome’s death reveals a striking divergence between Eastern and Western Christian traditions. In the Christian East, from the Syriac Epistle of Herod to Pilate to the elaborations of Symeon Metaphrastes and Nicephorus Callistus Xanthopoulus, the narrative of her death on the ice achieved rich rhetorical development. In the Greek tradition in particular, her death was construed as a carefully shaped typological inversion of John the Baptist’s martyrdom: the dancer who had demanded the prophet’s head is herself decapitated “not by iron but by ice,” her fate staged as a symmetrical act of divine retribution. Syriac, Coptic, and Arabic retellings tend toward cumulative intensification, multiplying punitive motifs to achieve a more expansive and dramatic vision of divine vengeance. Despite these differences in narrative approaches, Eastern authors consistently treat her death as a theologically charged act of retribution that retrospectively interpret her role in the execution of John the Baptist.

In the medieval West, by contrast, the legend never attained comparable narrative or rhetorical density: apart from the Latin Epistle of Herod to Pilate, it survives only in sparse, compressed notices invoking the dancer’s drowning or is redirected towards the motif of the earth swallowing the wicked alive. The early modern period, however, marks a decisive shift in the legend’s reception in the West. With the advent of print and the translation of Byzantine historiographical and hagiographical works into Latin, especially those of Symeon Metaphrastes and Nicephorus Callistus Xanthopoulus, the fully elaborated Greek version of the legend entered Western Europe in a stable and widely reproducible form. This phase of transmission underscores the extent to which the formation of “tradition” depends not only on narrative content but also on material conditions of circulation: translation, print, and patterns of readership reshaped the West’s engagement with the legend. Its rapid reception in early modern devotional and didactic literature suggests that its earlier marginality in the medieval West resulted less from any intrinsic lack of narrative appeal than from the contingencies of transmission. Once available, the legend proved highly adaptable, moving from learned Latin compilations into vernacular piety and, ultimately, into the literary culture of the late nineteenth century.

Yet the significance of this trajectory extends beyond the history of a single legend. The case of Salome’s icy death illustrates the dynamics by which apocryphal traditions emerge, circulate, and acquire meaning within different Christian cultures. It demonstrates that apocrypha are not merely marginal in relation to canonical texts but can become sites of narrative experimentation, where biblical figures can be reimagined and moralized in response to local rhetorical and theological priorities.

More broadly, this case contributes to our understanding of how religious imaginaries are formed and transformed. The figure of Salome, only briefly and anonymously sketched in the Gospels, becomes in these traditions the focal point of an evolving narrative economy of guilt, punishment, and spectacle. Her death—whether by ice, water, sword, or earth—is repeatedly reconfigured to express theological concerns about justice, remembrance, and retribution. In this respect, the legend may also be read against a wider scriptural logic of narrative completion: just as the figure of Judas, whose betrayal anticipates the death of Christ, is brought to a morally charged end in the Gospel tradition through his self-inflicted death, so is the dancer who caused the beheading of John the Baptist brought, in apocryphal elaboration, to a correspondingly exemplary punishment. The analogy points to a shared interpretive impulse to supply closure to acts of transgression through narratively appropriate forms of retribution.

Finally, the above account of the legend’s dissemination illustrates how collective imagination operates at the intersection of text, performance, and belief: it is through the accumulation, variation, and recombination of narrative motifs that a minor biblical figure acquires a vivid and enduring afterlife. The history of Salome’s icy death, therefore, offers more than a case study in the transmission of an apocryphal legend. It provides a window into the processes by which Christian communities, across languages and centuries, expanded the narrative world of Scripture, negotiated its moral implications, and inscribed those interpretations into cultural memory.