Ancient graffiti – defined here as an ‘informally-carved text placed secondarily on either living rock, or an existing structure or monument’ – have traditionally attracted attention from scholars for allowing unmediated access to the personal expressions of ancient individuals.Footnote 1 For this view, late antique and Byzantine graffiti present a problem. The informal texts from the fourth to fifteenth centuries are formulaic and repetitive, with the same simple prayers βοήθει ‘help’, ἐλέησον ‘have mercy’, and μνήσθητι ‘remember’ appearing time and again throughout the eastern Mediterranean.Footnote 2 A key question emerges: if personal expression was filtered through socially acceptable formulae, does this material still reveal individual experience? In recent years, scholars have contended with this issue by emphasizing what late antique and Byzantine graffiti can teach us about community practice, rather than individual identity. This is a view espoused throughout the 2021 volume Cultic Graffiti in the Late Antique Mediterranean and Beyond. In their introduction, the editors contend that ‘our graffiti writers were not seeking to express individuality, rather they wanted to join an established community’.Footnote 3 Marlena Whiting emphasizes that through using the same formulae as those who had written their name on architecture before them, a graffito author would contribute to their community.Footnote 4 The general trend is the assumption that whilst late antique and Byzantine graffiti are important evidence, they do not substantially reveal individual expression.
In this article, however, I want to examine the issue from a different angle. Whilst Greek graffiti, and particularly Greek prayer graffiti, from the late antique and Byzantine eras are formulaic, several notable exceptions exist in which we are able to glean an insight into the mindset of the authors. At these sites, the authors often disregard or alter set religious phrases to make explicit the help they need from God. The three sites discussed here, each from within the borders of the modern Greek state, are a prison in Corinth and sites from two Cycladic islands (Syros and Tinos). These three locations have been selected because their graffiti span from the fourth to the thirteenth century and employ many different epigraphic formulae. Furthermore, the graffiti at these sites share a unifying factor: the authors of the graffiti experienced overt or potential danger and expressed their desires for safety explicitly. It should be noted that these texts are not the only examples in which individual emotion can be read into graffiti, but they are some of the most illuminating, with a large corpus of texts worthy of discussion, and therefore have been chosen here to keep a tight focus.Footnote 5 Each site will be examined in context, noting adherence to and divergence from typical prayer formulae. Such decisions by the authors suggest information about their experiences.
Studying graffiti
Before launching into my case studies, it is necessary to comment on some of the methodological problems which accompany the study of graffiti, as well as establish the corpus of data from which I draw. The key methodological issue is dating. It is exceptionally rare for Greek-Byzantine graffiti to be inscribed with a date, meaning scholars are frequently reliant on archaeological context. Other factors such as palaeography can be used to try and date this material, but when graffiti is roughly scratched or painted, or simply poorly preserved, the palaeography can be uncertain. The three sites discussed in this article have all been dated to various periods between the fourth and thirteenth centuries; however, as I will show with a brief discussion of context and content, these dates are rarely secure.
Throughout this article, I will contextualize the case studies I discuss by comparing them to the wider late antique and Byzantine graffiti tradition. For this purpose, I will compare the graffiti to the holistic study from my PhD thesis, where I collected 1733 graffiti from Greece, Egypt, Syria, Jordan, Turkey, North Macedonia, and Israel/Palestine.Footnote 6 Of this material 920 inscriptions are both textual and legible, and it is this selection of material I refer to when I discuss percentages and the broader graffiti tradition. This data is not inclusive of later Byzantine traditions; however, a cursory examination suggests a continuation of the same practices.
Corinth: prayers from a Late Antique prison
The late antique graffiti at Corinth were discovered on paving stones in the north-west shops of the agora (Fig. 1), in a room adjacent to the central ‘boudroumi vault’. The paving stones appear to have been associated with modifications to the Roman structure in the late antique era.Footnote 7 A similar pavement slab bearing the same material and stylistic qualities was found near the north-west shops, featuring an acclamation of the emperors Justin II and Tiberios, which may allow us to narrow down these modifications to the sixth century, assuming the imperial acclamation was added during or shortly after renovations.Footnote 8 The graffiti were discovered in 1901 and first published in 1931 by Benjamin D. Meritt, but further examination of the inscriptions in context has been prevented by their removal from the north-west shops to storage.Footnote 9 That this space functioned as a prison can be ascertained from the content of the graffiti, including one reference to βουκελλάριοι ‘guards’ and others to different forms of captivity or restraint.Footnote 10 Various dates have been proposed for the graffiti, from the fourth to tenth centuries; though if we accept a sixth century date for the renovations this can be narrowed to the sixth to tenth centuries.Footnote 11 At first glance, the extant graffiti from Corinth adhere to the prayer formulae we would expect to see in late antique graffiti. There are two examples of the ‘Lord, help’ formula or its variants, for example, ‘Lord Jesus Christ, help your servant Kosmas. Amen’.Footnote 12 These inscriptions firmly root the Corinth graffiti within the broader epigraphic landscape of the late antique world.
The Corinthian agora in late antiquity, with the ‘boudroumi’ vault indicated in red. Source: Scranton, Medieval Architecture, plan IV.
However, not all prayer texts contain the expected formulae. Some inscriptions instead refer to the explicit dangers faced by those incarcerated at Corinth. The language used was personal and in relation to a specific set of circumstances felt by each prisoner. One text, for example, reads ‘may the fortune triumph … do not judge unfairly (παραδίκασον) between me and my enemies.’Footnote 13 Παραδικάζω (which Geoffrey Lampe translates as ‘pervert justice’, may also be rendered ‘judge unfairly’) is not seen in graffiti at other sites in the eastern Mediterranean in this period.Footnote 14 There is no particular reason that the perversion of justice would be preeminent in the minds of the average pilgrim or urban inhabitant. However, the unfair judgement ‘between me and my enemies’ was a pressing concern for this prisoner – and they necessarily addressed God expressing this concern. Although the verb παραδικάζω does not appear in other graffiti from Corinth, the imperative of ἀπόλλυμι (or ἀπολλύω) ‘destroy’ appears twice. One text wishes destruction on an unnamed enemy, whilst also referencing ‘this place’ (as reconstructed by the editor) in which they are imprisoned, highlighting their distress at their incarceration.Footnote 15 Their enemy was likely the individual responsible for their imprisonment, probably a certain ‘Marinos’ who is referenced in two other inscriptions. One text also employed the term ‘destroy’ to request that God bring harm to Marinos. It read ‘God of Justice who judges rightly, strike quickly against the Hellenes … and destroy the hated … ]eros and Marinos the sons of Ioannes the barber.’Footnote 16 The epithet ὁ Θεὸς δίκης, ‘God of Justice’ emphasizes that God is invoked in his role as divine arbiter of right and wrong. The term Hellenes is also noteworthy: in late antiquity this term often took on a pejorative meaning against pagans, and thus its appearance here may indicate that the author's imprisonment was underscored by a religious conflict.Footnote 17 (The term was fluid, however, and the appearance may also indicate the author did not consider themselves as having a Hellenic identity.) Another graffito also requests God's intervention in the conflict between the author, the incarcerated Petrounia, and Marinos, ‘Theotokos, punish Marinos who has thrown us here and release (with) us Petrounia.’Footnote 18 One inscription requested Leonianos (who is similarly blamed for the authors imprisonment) is destroyed.Footnote 19 Another inscription reads: ‘make them die a bad death’ and may have referred to those who imprisoned the author.Footnote 20 Throughout this material, texts also explicitly request the author's liberation.Footnote 21 In these dire situations, it is apparent that the Corinthian prisoners used not only common formulae, but developed their own in response to imminent danger.
Also notable is how the graffiti at Corinth inverted and adapted common late antique prayer formulae and acclamations to fit the circumstances of the author. This is particularly overt in one example: ‘may the fortune of those suffering in this lawless place triumph! Lord, have no mercy on (those who) threw me here’.Footnote 22 Here, the common prayer ‘Lord, have mercy’ was inverted to wish harm on those who imprisoned the author.Footnote 23 The phrase ‘Lord, have no mercy’ indicated the author had a familiarity with the language of inscribed prayer but did not believe it suited their present needs. Rather, they found a way to utilize this formula to relay their desired message to God. The same phenomena can be recognized in the appropriation of the phrase νικᾷ ἡ τύχη ‘may the fortune triumph’, which appears in four inscriptions in the prison at Corinth.Footnote 24 Νικᾷ ἡ τύχη was a common phrase in the spoken and epigraphic language of the eastern Roman empire. It appeared in the Book of Ceremonies, spoken during the proclamation of Anastasios I ‘may the fortune of the Romans triumph’.Footnote 25 A variant of this phrase, ‘may the faith triumph’ was also spoken during ceremonies, by the Green and Blue circus factions.Footnote 26 The individuals who used this term during public ceremonies also carved the phrase in the entertainment venues of the late antique city: for example, a graffito from the theatre of Aphrodisias, which reads ‘may the fortune of the Greens, and the Green mime, triumph’.Footnote 27 This acclamation, which would have been familiar to the urban inhabitants at Corinth, was appropriated by the prisoners to invoke their own fortune in their treatment by judicial authorities.
The graffiti from Corinth provide a key insight into the personal expression of late antique people.Footnote 28 The common formulae which existed in the late antique Greek cultural context were used by those incarcerated in order to request divine aid, however they were not used in isolation. The appearance of specific requests is indicative of the fact that when the author of a graffito had a request of God, they deviated from the use of normal epigraphic prayer. Requests for the destruction of those who imprisoned the authors, and fair judgement for the incarcerated, were distinctive of the situation of the imprisoned, and thus their prayers contain similarly unique terminology. The extant graffiti all display the authors’ awareness of the common epigraphic formulae of religious and urban environments, and their willingness to appropriate these formulae to wish harm on their enemies or call for their own vindication.
Grammata Bay, Syros: travellers and the dangers of the sea
Grammata Bay, on Syros in the Cyclades,Footnote 29 is a natural harbour which functioned as a port in antiquity.Footnote 30 Visitors to the island carved their names and prayers upon the rocks of the harbour, with graffiti dating from the third century through to the medieval era.Footnote 31 The authors of the Christian graffiti were travellers visiting Syros, either as a final destination or as part of a longer journey. The majority of the visitors appear to have been sailors or ship's captains travelling for economic purposes, although some authors identified themselves by their ecclesiastical and professional titles: reader, deacon, goldsmith.Footnote 32 Others used military and naval titles such as optio, a position within the Roman and late Roman army which had several associations.Footnote 33 Visitors to Syros appear to have originated from a diverse range of areas, but two main groups emerge. The first is travellers from within the Cyclades, with fifteen texts indicating the author originated from another island.Footnote 34 The most common of these is Naxos, the largest Cycladic island, named as a place of origin in four inscriptions.Footnote 35 Smaller Cycladic islands are also referenced, such as Gyaros.Footnote 36 The second group of travellers are those coming from further distances, including Asia Minor and the near east. One text names Ephesos as a place of origin, whilst others name Tyre and Bithynia (Fig. 2).Footnote 37 Georges Kiourtzian notes that the lack of female names in the inscriptions suggests travellers were unlikely to be motivated by pilgrimage.Footnote 38 The demographics of travellers generally allows us to reconstruct the visitors to Syros as men motivated to travel for civic, military, or economic reasons.
Map depicting origin of travellers to Syros. Source: Nowakowski, Pilgrims and Seafarers, fig. 9.4.
The Christian graffiti at Grammata Bay have typically been dated to late antiquity. The material, which is abundant and shows a close familiarity with Christian formulae, can likely be dated to a period of relatively high Christianization within the Roman empire. For these reasons, Kiourtzian suggests a dating of the fifth century onwards, with some as fourth-century in origin.Footnote 39 Kiourtzian generally provides a terminus ante quem of the year 649, arguing that the presence of the Arab fleet in the Cyclades would preclude shipping in the region.Footnote 40 This is with the exception of two inscriptions, which Kiourtzian suggests may be twelfth to thirteenth century. However, there is evidence that this reading should be broadened. For example, one graffito lists the naval title κένταρχος ‘kentarchos’, a middle Byzantine title, and Kiourtzian suggests this is the earliest reference to that position.Footnote 41 This only follows, however, if one accepts Kiourtzian's unpersuasive claim that 649 represents a definitive end date for travel within the Cyclades. Whilst the Arab fleet may have disturbed long-distance travel between some of the cities and regions featured in the graffiti (Tyre, Ephesos and Bithynia), there is little reason to think this prevented more local travel in the seventh century and beyond. Archaeological evidence from Naxos indicates that the Cycladic islands remained relatively well-connected to one another and the Greek mainland during the seventh to tenth centuries.Footnote 42 Consequently, even if long distance travel declined, we should not assume a complete abandonment of inter-island travel in the middle Byzantine Cyclades. We might therefore, accept a dating of the graffiti at Grammata as between the fourth and twelfth or thirteenth centuries, with the earlier period being that of greatest epigraphic production.
Again, it is important to read the graffiti from Syros within the context of late antique and Byzantine epigraphic prayer formulae. Requests for divine aid are among the most common Christian inscriptions at Grammata Bay. In total, of ninety-one inscriptions, forty-seven feature either ‘Lord, help’, ‘Christ, help’, or variants. As Grammata was not an urban location with a formal epigraphic tradition, this use of common formulae is revealing: it indicates that early Byzantine individuals were acutely aware of the terminologies which dominated their informal epigraphy, and chose to replicate this language themselves.
However, as with Corinth, the graffiti from Syros contain language which references the danger faced by the authors. In the case of Syros, however, it was not the risks associated with imprisonment and trial, but those associated with sea travel. The sea was represented as a dangerous environment in classical, late antique and Byzantine culture. The threat of storms, piracy, and being cast adrift were ever present. Synesios of Cyrene recorded his own experience of being stranded at sea following a storm and the near-death experience it precipitated; he ended his letter advising his brother ‘may you never trust yourself to the sea’.Footnote 43 The same fear of harm whilst at sea is manifested in material culture such as pilgrimage tokens. A stamp for creating these tokens depicts Saint Isidore of Chios holding a ship (Fig. 3).Footnote 44 Another pilgrimage token depicting Saint Phokas, who was particularly revered by sailors, standing aboard a ship, complete with oars and a rudder, which acted as a placeholder for the ship a pilgrim would take on his or her journey, whilst the image of the saint invoked his presence and protection.Footnote 45 For the sailors travelling to Syros, it is therefore not surprising they expressed fear of sea travel in their prayers. The Cyclades suffers from strong winds and weather variability – particularly in the spring and summer when most shipping would have been taking place, meaning that a desire for protection from storms would have been particularly relevant.Footnote 46 Of inscriptions at the bay 37% carry explicit references to ships or crew.Footnote 47 One text prays for the safety of the ‘Georgios’ ship and Petros the captain with his sailing companions’.Footnote 48 Another two graffiti name the ships which travelled.Footnote 49
Stamp for the creation of a pilgrimage token, featuring Saint Isidore of Chios holding a ship. © The Walters Art Museum. https://art.thewalters.org/detail/22682/pilgrim-stamp-of-saint-isidore/. [Accessed 07.11.2023].
The desire for protection whilst at sea is made even more explicit through the use of the term εὔπλοια, ‘fair voyage’. Seven late antique graffiti at Grammata feature some such invocation. For example, a Christian inscription which finishes ‘grant to him good sailing, amen’.Footnote 50 Invocations for ‘good sailing’ were neither unique to Syros, nor a Christian innovation. According to the Packard Humanities Institute (PHI) database of Greek inscriptions, εὔπλοια appears in some form in 75 inscriptions (not accounting for spelling variations).Footnote 51 Twenty-seven originate in the Aegean islands; others are found along the mainland coastline. For example, of the fourteen inscriptions featuring which the PHI lists in the Peloponnese, thirteen are found on the island of Proti, and one is found at Kolonides on the Messenian coast.Footnote 52 The appearance of εὔπλοια in regions associated with the sea and sea-travel indicates its association with the cultural fear of the sea from this era. At Syros itself, the invocation for ‘good sailing’ appears in explicitly non-Christian texts, for example ‘Good sailing to Ioulianos of Milos the Serapis-lover, son of Artemisios’ highlights the definitive use of the phrase from an era which is certainly earlier than most, if not all, of the Christian invocations.Footnote 53 The continuation of the use of the ‘Good sailing’ formula, and references to ships and professions in the pre-Christian graffiti from the site into the late antique and Byzantine inscriptions, indicated a universal need of sailors to be protected from dangers of the sea.
Of special note is the use of the σῶσον ‘save’ prayer formula at Grammata. Although the prayer is recognized in multiple contexts in the late antique and Byzantine worlds (both textual and material) it is surprisingly uncommon in graffiti, appearing just thirty-five times, or in roughly 4% of all textual graffiti.Footnote 54 Notably, twenty-three of these appearances are from Syros. We might therefore wonder if requests to be saved carried a special meaning amongst the sailors stopping at this island. I posit that the specific choice of the term ‘save’ was due to the need for physical salvation from the dangers of the sea and sailing, something evident in in those texts which mention ships. For example, ‘Lord, save the ship of Georgios and Petros …’.Footnote 55 Furthermore, whilst none of the pre-Christian inscriptions make use of the term βοήθει, there is a possible pre-Christian use of σῶσον, although it should be noted this text has been reconstructed.Footnote 56 The use of the term suggests it was not exclusive to spiritual salvation, but was also associated with physical needs. It may be that Christian authors had few qualms about adopting this to request physical salvation, especially as it aligned neatly with pre-existing Christian prayer formulae. The use of σῶσον to request physical salvation is mirrored at other sites in the eastern Mediterranean, where the author of a graffito would have been endangered whilst travelling. From the late antique Greek graffiti along pilgrimage routes to St Catherine's monastery in the Sinai desert, σῶσον appears seven times, including one prayer which specifically mentions travel, ‘Lord, save your servant Theophilos and those who belong to him, and his travelling companions. Amen’.Footnote 57 A certain correlation exists, therefore, between graffiti written by those on dangerous journeys, and those who selected the ‘save’ formula for their informal prayers.
The prayer formulae found in graffiti at Syros were carefully chosen by the authors, who altered their traditional prayers to accommodate the dangers they experienced during travel by sea in the Cyclades. Prayers were augmented with references to ships, sailors, and cargo in order to ensure God's recognition of the dangers of sea. The term ‘save’ was a part of a culturally defined roster of divine requests chosen by the sailors at Syros to ensure the safe completion of their journeys. It is important to note that these authors were not divorced from the ‘formulaic’ graffiti of the day, and made use of prayers such as ‘Lord, help’ frequently, whilst even prayers such as ‘save’ were not exclusively associated with safe travel. Rather, when selecting from this roster of prayers, they often chose one which they believed was most appropriate to their current needs.
The cave chapel of Saint Stephen, Tinos: selecting prayer formulae in the middle Byzantine era
The final site I will examine is the Cave Chapel of St Stephen on Tinos. The cave is located on the southern tip of the island, close to the Hellenistic sanctuary of Poseidon and the modern village of Gastria.Footnote 58 Denis Feissel discovered forty-eight Christian graffiti at the cave, which he published in 1980, accompanied by two additional inscriptions originally recorded by Hubert Demoulin. In 2000, Kiourtzian re-edited three of the inscriptions published by Feissel and published an additional three fragmentary graffiti from the cave.Footnote 59 The graffiti within the cave can most likely be ascribed to the middle Byzantine period due to textual similarities with the graffiti in the Parthenon, which can be securely dated to the late seventh to thirteenth centuries.Footnote 60 Nevertheless, despite the plethora of inscriptional material, the purpose of the cave remains relatively obscure to us. There is no extant evidence of a systematic decorative programme, which we would expect from a church or chapel.Footnote 61 It is possible that, as with Syros, the cave did not have a devotional function and merely provided a dry and convenient space for sailors to shelter.Footnote 62 At the same time, however, whilst the Christian graffiti at Syros do not consistently invoke the aid of any specific religious figure besides ‘Lord’ or ‘Christ’ (excluding one reference to St Phokas) twelve graffiti at Tinos acclaim Stephen Protomartyr, suggesting that unlike Syros, the cave at Tinos did bear some association with a Christian saint.Footnote 63
Additionally, a modern chapel dedicated to Saint Stephen is placed nearby, and may therefore represent a possible continuation of this tradition, although it may also be the result of coincidence or the modern rediscovery of the Byzantine texts.Footnote 64 Thus it is probable the cave did have a cultic function and local association with the saint, and its visitors should therefore be seen as pilgrims.Footnote 65 Three graffiti in the cave list women as the primary supplicant, further supporting this suggestion.Footnote 66 Pilgrims to the cave appear to represent both local and foreign supplicants. Two inscriptions explicitly identify the author as from Tinos.Footnote 67 Other inscriptions indicate a foreign origin, with the authors identified as originating at several locations from the Aegean and Asia Minor. For example, two graffiti name Timotheos, Bishop of Knidos.Footnote 68 The date of Timotheos’ visit to the cave is uncertain, with both Feissel and Kiourtzian favouring a pre-seventh century date, on the assumption that travel across the Aegean would have been impossible in later periods due to the Arab navy, and due to the apparent disappearance of the title Bishop of Knidos in the eighth century.Footnote 69 As has already been discussed, Arab presence is not a strong enough argument to suggest travel could not have occurred, additionally it appears that the title of Bishop of Knidos re-emerged in the tenth century.Footnote 70 Accordingly, a middle Byzantine date for this graffito is still possible. Another author travelling from Asia Minor was ‘Eustathios of Paphlagonia’.Footnote 71 Another originated from the Greek mainland: ‘Ioannes of Athens’, perhaps a goldsmith.Footnote 72 Even if these travellers did not journey to Tinos on pilgrimage, they chose to visit the chapel to partake in the cult of St Stephen whilst on the island.
As with Corinth and Syros, the graffiti at Tinos largely align with the epigraphic traditions we would expect to see in the Byzantine era. The most common prayer inscribed in the cave chapel is ‘Lord, help’, which appears in 72% of graffiti at the site.Footnote 73 There are also many local characteristics of the prayers at Tinos. Eleven inscriptions finish with the term ‘amen’, which is uncommon in Byzantine graffiti, but not unprecedented. More notable however, are the six inscriptions which also include the Greek translation of ‘amen’, γένοιτο or ‘so be it’. Ἀμήν, γένοιτο appears in the graffiti at three other known sites in the eastern Mediterranean, in one graffito at Panormon in Crete, one text from Ephesos in Asia Minor and in four texts at the Parthenon in Athens.Footnote 74 As we know from the inscription of Ioannes the Athenian that travel between Athens and the cave chapel of St Stephen occurred, the use of ἀμήν γένοιτο should probably be read as a regional variation in graffiti, specific to Athens and those places in frequent contact.Footnote 75 Regardless of this local variation however, the frequency of the ‘Lord, help’ formulae associate these graffiti firmly with the typical nature of epigraphic Byzantine prayer.
As with Syros, however, we see that a selection of graffiti includes reference to the dangers of the Cyclades. The cave is only meters from the shore, and those who worshipped there were either visitors who travelled to the island by ship or locals who would have been regular witnesses to the threats storm surges and flooding could cause. One text explicitly references the dangers of the sea, the author requesting that God ‘save him, Lord, in the sea’.Footnote 76 Another pilgrim similarly referenced the dangers caused by storms in his prayer ‘Lord, help your servant Sedorros… release him from the surge of the sea.’Footnote 77
Interestingly, by contrast with Syros, we do get a selection of texts which reference theological concerns, a factor which contributes to our reading of the Tinos cave as a cult environment or consecrated shrine, rather than merely a dry space for sailors to shelter. One Papaglegores requested that he be free ‘from his sins, and free him from the indistinguishable fire’.Footnote 78 Similarly, a supplicant named Anna asked to be released from her sins, whilst two pilgrims, Gabril and Eirene, asked for the ‘requests of their heart’.Footnote 79 One inscription, which does not name its author, does not feature any prayer but instead relates to scripture. The very unusual invocation refers to Matthew 27:51, and is paralleled not in informal inscriptions but in formal church epigraphy.Footnote 80 Within this context we should therefore pay especial note to those inscriptions which reference the sea. They reveal that within this cultic environment, the fear of physical harm led to individuals expressing their desire for salvation from the sea.
A different category of graffiti in the cave chapel of St Stephen also illuminates the relationship between dangerous travel and prayer in graffiti. Seven graffiti feature either all, or part, of the phrase σκέπε, φρούρει καὶ φύλαττε ‘shelter, protect, and guard’, e.g. ‘Saint Stephen, shelter, protect and guard your servant Basilios, the lowest Archon, and his wife and his children’.Footnote 81 The use of this formula at Tinos can be explained in multiple ways. Variants of ‘shelter, protect and guard’ appear multiple times in the graffiti at the Athenian Parthenon, e.g. a graffito reading ‘Lord, our God, shelter and protect your servant Grigorios, Archivist, Amen’.Footnote 82 The terms ‘shelter and guard’ also appear in a formal inscription at the Parthenon, indicating the use of this phrase across multiple forms of inscription.Footnote 83. This prayer is also attested on another island of the Cyclades, in an inscription from Naxos.Footnote 84As with ἀμὴν γένοιτο, it is possible that ‘shelter, protect and guard’ was a localized Athenian and Cycladic prayer dispersed through travel to Tinos. However, I would like to propose an alternative theory which may have contributed to its adoption and use at Tinos. Although the formula is not recognized in any other Byzantine graffiti outside of Athens, the phrase is common on middle Byzantine seals and eulogia.Footnote 85 A tenth-century seal, currently in the Dumbarton Oaks collection, reads ‘Cross, shelter, protect and guard your servant Konstantinos the priest’.Footnote 86 Another example reads ‘Theotokos, guard and protect the most venerable Soteirios.’Footnote 87
What then might have motivated pilgrims and other supplicants at the chapel to adopt a phrase from seals when carving their graffiti? The answer may lie in the similar actions taken by seals and eulogia, and the individual traveller at Tinos. Eulogia were carried from a pilgrimage site to the supplicant's home, regardless of distance. Additionally, seals were responsible for the safe deliverance of a letter or object. The desire a seal would ensure letters arrived at their destination unaltered is evidenced in the inscription of one seal, which uses the terms ‘guard’ and ‘protect’ to refer to the seals purpose ‘I am the seal of Leo Kouboukleisios, which both guards and protects my things’.Footnote 88 Another seal from the eleventh or twelfth century reads ‘The seal (is) guarding the words of Theophylaktos’.Footnote 89 It is possible the ‘shelter, protect and guard’ formulae were associated not only with local means of expression, but specifically with travel and the safe transferal of an individual or object from their place of origin to their destination. Thus, the supplicants who either arrived at Tinos by sea, or were locals who engaged in activities involving the sea, chose this phrase in order to ensure their physical safety. Once again, we see how the environment of the inscription and context of the author led to the adoption of uncommon yet meaningful formulae.
Before concluding, it is necessary to make a few comments on the palaeography of these graffiti. Of the seven inscriptions which request some variation of ‘shelter, protect and guard’, four have notable similarities in the handwriting, particularly in the form of ἅγιε ‘Saint’.Footnote 90 All four of these graffiti open with an uncial alpha, followed by a majuscule gamma with a small vertical strike attached to the horizontal bar (Figs 4–5). The similarities in these inscriptions would hint at the same author inscribing each one; and as all four name different individuals, perhaps one author was paid to write these texts. If this was the case, then this author may have been an individual named Agathon, whose name appears in one text which also references ‘the writer of this’.Footnote 91 In the event that a writer, such as Agathon, acted as an intermediary, this does not discredit the importance of the formulae for understanding personal experience. The individual who commissioned or requested the text may very likely have had a choice of which formulae would be written.
‘Shelter, protect and guard’ graffiti from Tinos. Source: Feissel, Ténos, nos. 5 (top), 11 (center), and 12 (bottom).
Graffiti from Tinos. Source: Feissel, Ténos, nos. 13 (top), 14 (center) and 15 (bottom).
For example, it is worth noting that it is not only these four texts which are written in the same hand, but four additional texts feature the distinctive ἅγιε (Figs 5–6).Footnote 92 However, these texts each feature the prayer ‘help’, rather than any variant of ‘shelter, protect and guard’.Footnote 93 The use of a different phrase in these four inscriptions would suggest that perhaps, even if the texts were commissioned, the supplicant perhaps still had the opportunity to select the prayer they wished to be inscribed on their behalf, and selected the one they believed would grant them salvation from the sea. This is a suggestion supported by the appearance of two graffiti written in unique hands which were also found in the cave, which request protection for the author. One graffito written by, or on behalf of, an individual named Amanos, requests God ‘guard’ him.Footnote 94 This graffito was certainly written by a different author than Agathon, as is clear through the form of the alpha. In the graffito of Amanos, the alpha is majuscule, with a deep ‘V’ shape forming the cross-bar (Fig. 7). Similarly, another graffito (the personal name written here has been lost) utilizes the ‘shelter, protect and guard’ formula.Footnote 95 Although this graffito does share palaeographic similarities with the Agathon texts, such as the uncial alpha and the strike on the gamma, the graffito also makes use of the pi/rho ligature for πρωτομάρτυρος, which do not appear in the Agathon examples (Fig. 8). In all these inscriptions, regardless of whether they were inscribed by the supplicant or an intermediary, the use of the rare ‘shelter, protect and guard’ formula was a deliberate choice, motivated by the environment in which the texts were written.
Graffiti from Tinos with an uncial alpha and gamma with a short line. Source: Feissel, Ténos, nos. 4 (left) and 27 (right).
Graffiti from Tinos requesting protection. Source: Feissel, Ténos, no. 10.
Graffiti from Tinos requesting protection. Source: Feissel, Ténos, no. 1.
As at Syros, inhabitants and visitors to Tinos were acutely aware of the dangers they faced in the island environment, and thus emphasized their desire for protection in their prayers. Supplicants carved requests for their protection from the sea, whilst also emphasising their position in relation to the sea, such as their occupation as sailors. In some cases, the authors used a local formula which was also associated with the safe delivery of letters to request their own safe delivery when travelling. The use of such a formula highlights the breadth of the Byzantine epigraphic tradition and Byzantine people's understanding and familiarity with this breadth. When inscribing a text, they did not always blindly adopt common phrases but selected phrases which related to their current situation and needs.
Conclusions
Christian graffiti in late antiquity and Byzantium has traditionally been characterized as formulaic, not providing the insight into the personal experiences and emotions we expect to read in graffiti from earlier periods. This formulaic nature has invited greater attention in recent years, however the underlying belief that the repeating phrases associated with Christian inscriptions belie a detailed understanding of individual experiences prevails. Late antique graffiti certainly were a repetitive medium, with Christian inscriptions frequently featuring prayers, invocations and acclamations from a small selection of phrases. This article has not sought to disprove the formulaic nature of late antique graffiti, but rather to question the implication we cannot read the authors experiences in the text. When Byzantine peoples were placed in dangerous scenarios, they wrote prayers relevant to these scenarios.
At Corinth, prisoners expressed their desire for freedom, and vengeance on those who had incarcerated them. These prisoners both used unparalleled formulae related to their specific situation (including calls for justice and the destruction of their enemies) and drew formulae from a cultural catalogue of epigraphic expressions. Through the appropriation and altering of phrases such as ‘Lord, have mercy’ (becoming ‘have no mercy’) and ‘may the fortune triumph’ individuals were able to express their immediate needs to God within an understood linguistic framework. Similarly, at Grammata Bay in Syros, sailors and other travellers continued a long tradition of carving informal prayers to their deity, in the hope they would receive salvation. Within these prayers they identified both the individuals and the ships which required salvation, as well as textually requesting their primary need: ‘good sailing’. Furthermore, the prevalence of the ‘Lord, save’ formulae appears to have been associated with the physical danger from which the author required salvation. At Tinos we see a middle Byzantine era continuation of these prayer traditions. Authors from the island landscape and travellers between islands expressed not only their fears of travel by water, but also specifically adopted formulae associated with the safe delivery of an item to its destination, the seal formula ‘shelter protect and guard’. We thus can recognize the connection between the intensity of a person's emotions (e.g., fear for their safety) and the graffiti they carve.
What this article does not seek to argue is that Byzantine graffiti were not formulaic. In fact, as I have shown, Byzantine graffiti was overwhelmingly formulaic, with the vast majority of texts adhering to specific formulae, be they ‘help’, ‘have mercy’, ‘remember’, ‘save’, or ‘shelter, protect and guard’. However, these formulae were chosen with purpose. When an author felt an acute sense of danger, such as the threat of drowning at sea, or the consequences of an unjust incarceration, the author accounted for this by selecting formulae which were pertinent to the situations in which they found themselves. These conclusions appear across all aspects of the graffiti tradition, in both secular and cultic spaces, when the supplicant was the author, or when they may have acted through an intermediary. Communities selected the phrases most relevant to their own salvation, and employed these formulae accordingly. Going forward, the formulae chosen must be examined carefully in order that we may better reconstruct the emotions and experiences of individuals in a post-classical world.
Rachael Helen Banes is a postdoctoral researcher at the Austrian Academy of Sciences on the Epigraphies of Pious Travel project. Rachael completed her PhD at the University of Birmingham in 2023, on the subject of late antique graffiti in the eastern Mediterranean. She is currently researching Greek pilgrimage graffiti throughout the Byzantine period. Her research interests include the epigraphy, material culture, and urban environment of the late antique and Byzantine era.
