Erhalten Sie Zugang zu diesem und mehr als 300000 Büchern ab EUR 5,99 monatlich.
- Herausgeber: Narr Francke Attempto Verlag
- Kategorie: Geisteswissenschaft
- Serie: Sprachvergleich
- Sprache: Englisch
The edited volume presents the rst comprehensive corpus of performative adjuration formulae collated from Greek and Latin epigraphical sources. The original texts-for the most part artefacts connected with magico-religious beliefs and practices of their users-are all translated into English and accompanied by a philological and socio-religious commentary. The international team of three specialists adopts a synoptic approach that tracks various classes of epigraphic documents to analyse permutations and developments in the syntactic structure of the adjuration formula, and its pragmatic function. This major study of the adjuration formula in Antiquity and its continued tradition in the Middle Ages will be of interest not only to the scholars of these linguistic traditions, but also to researchers working in the elds of Religious Studies, Ancient History, Theology, and Archaeology.
Sie lesen das E-Book in den Legimi-Apps auf:
Seitenzahl: 747
Veröffentlichungsjahr: 2025
Das E-Book (TTS) können Sie hören im Abo „Legimi Premium” in Legimi-Apps auf:
Ähnliche
Juraj Franek / Daniela Urbanová / Ulrike Ehmig
Performative Adjuration Formula in Greek and Latin Inscriptions
A Survey of Amulets, Curse Tablets, and Funerary Monuments
This work was supported by the Czech Science Foundation (GA21-06319S) and the Masaryk University, Faculty of Arts. Daniela Urbanová kindly acknowledges the support of the Commission for Ancient History and Epigraphy in Munich (Deutsches Archaologisches Institut), and especially its Vice Director Prof. Dr. Rudolf Haensch, for the opportunity to work on the paper in August 2023 and September 2024. We would like to thank T. Hlavička for preparation of the map.
DOI: https://doi.org/10.24053/9783823395461
© 2025 • Narr Francke Attempto Verlag GmbH + Co. KGDischingerweg 5 • D-72070 Tübingen
Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen.
Alle Informationen in diesem Buch wurden mit großer Sorgfalt erstellt. Fehler können dennoch nicht völlig ausgeschlossen werden. Weder Verlag noch Autor:innen oder Herausgeber:innen übernehmen deshalb eine Gewährleistung für die Korrektheit des Inhaltes und haften nicht für fehlerhafte Angaben und deren Folgen. Diese Publikation enthält gegebenenfalls Links zu externen Inhalten Dritter, auf die weder Verlag noch Autor:innen oder Herausgeber:innen Einfluss haben. Für die Inhalte der verlinkten Seiten sind stets die jeweiligen Anbieter oder Betreibenden der Seiten verantwortlich.
Internet: www.narr.deeMail: [email protected]
ISSN 2569-2275
ISBN 978-3-8233-8546-2 (Print)
ISBN 978-3-8233-0520-0 (ePub)
Contents
1Introduction
Over the past three decades, the study of epigraphic documents reflecting the religious and magical practices of the ancient Mediterranean peoples has experienced a major surge in scholarly interest. Important new studies and editions of Greek and Latin curse tablets have appeared;1 inscribed magical amulets of variegated types and material supports, such as papyrus or gold, silver, bronze, and copper lamellae, have been collated and edited in a series of comprehensive volumes;2 new important editions of funerary inscriptions and valuable studies of tituli sepulcrales are steadily being published.3
In addition to the works that have already seen the light of the day, the upcoming years are ripe with multiple expanded re-editions of classical works and brand-new publications. At the time of writing, the standard edition of Greek Magical Papyri4 and their English translation5 is being superseded by a new bilingual edition curated by Christopher A. Faraone and Sofía Torallas Tovar at the University of Chicago;6 the corpus of Greek Attic defixiones,7 well over a century old, has been replaced by the long-awaited second edition by Jaime Curbera as part of the Inscriptiones Graecae, published by the Berlin-Brandenburgische Akademie der Wissenschaften;8 a major new corpus of Sicilian inscriptions is being put together by a team led by Jonathan Prag at the University of Oxford.9 Brand-new editions of previously unpublished texts are also in preparation, such as Magica Levantina, a collection of curse tablets from the Promontory Palace in Caesarea Maritima, edited by Robert Daniel and Alexander Hollman at the Universität zu Köln; and an edition of defixiones from Uley, courtesy of Roger Tomlin at the University of Oxford. Large-scale, multi-volume editions that focus on specific geographical regions and collate all epigraphic texts irrespective of language, are also being steadily published, with some projects approaching finalization.10
Thanks to the advent of digital humanities, it is now possible to access the majority of these texts, both old and new, via different digital platforms. To include just a handful of examples, Thesaurus Defixionum (olim Otto von Guericke Universität Magdeburg and reestablished the University of Hamburg under the guidance of Werner Rieß and Sara Chiarini), collated most preserved Greek and Latin curse tablets; Inscriptiones Christianae Graecae (Humboldt-Universität zu Berlin) covered early Christian funerary inscriptions from Greece and Asia Minor; the last volume of the important series Monumenta Asiae Minoris Antiqua has been published exclusively in a digital format (University of Oxford).
All-purpose epigraphical resources of the likes of Inscriptiones Graecae (Berlin-Brandenburgische Akademie der Wissenschaften), Searchable Greek Inscriptions (Cornell University; Ohio State University), Supplementum Epigraphicum Graecum (Brill), Roman Inscriptions of Britain (University of Oxford), Epigraphik-Datenbank Clauss-Slaby (Katholische Universität Eichstätt-Ingolstadt), Epigraphic Database Heidelberg (Universität Heidelberg), Trismegistos (Universities of Leuven and Cologne), Computerized Historical Linguistic Database of Latin Inscriptions of the Imperial Age (ELTE, Budapest), Inscriptions of Israel and Palestine (Brown University), and several other databases are substantially facilitating access and search capabilities in the field of ancient epigraphy. In recent years, a major project funded by the European Commission, The Europeana Network of Ancient Greek and Latin Epigraphy, has started laying the groundwork for an integration of isolated databases into a meta-database collating all available material.
The double impact of a flurry of new editions and studies published in recent decades and the ever-increasing availability of these texts in the digital space create research opportunities that were not available even a decade ago. While scholarly interest in the performative and formulaic language of religious and magical epigraphic texts is certainly not nonexistent,11 it has been traditionally marginalized in favour of more ‘urgent’ tasks, such as the production of new editions and linguistic studies focusing on dialectal variety, sound changes, morphology, ‘late’ or ‘vulgar’ language, and cursing formulae (esp. in the case of Latin).12 With respect to the focus of our monograph, performative adjuration formula, there is a significant disconnect between the numerous scholarly works dealing with oaths, exorcisms and adjurations in literary sources, especially those connected with Jesus and Early Christianity,13 and the limited attention that has been paid to the study of the very same elements in epigraphic documents.14
No comprehensive survey of performative formulae across different types of epigraphical texts written in Greek and Latin has been published before now. As such, our primary aim has been to produce a compact, single-volume monograph focused solely on the performative use of adjuration formula across all types of Greek and Latin epigraphical documents, including funerary inscriptions, gems, lamellae, and papyri with spells of both apotropaic and aggressive magic. Adjuration formula has been part and parcel of ancient Mediterranean religious and magical discourse, its internal structure remains readily identifiable in both Greek and Latin linguistic tradition across centuries of use, and a synoptic, multi-lingual approach that tracks all classes of epigraphic documents makes it possible to demonstrate both its syntactic structure and the pragmatic function.
1.1Methodology
The corpus of epigraphical documents collated in this survey involves 171 entries in total, of which 119 are inscribed in Greek and 52 in Latin. The Greek material is not only more numerous, but also often more varied, with lengthier and more elaborated texts than its Latin counterparts. Performative adjuration formula has been attested almost exclusively in three types of documents: curse tablets (49 instances in total, with 36 Greek and 13 Latin), amulets (48 instances in total, 34 Greek and 14 Latin), and funerary inscriptions (71 instances in total, 49 Greek and 22 Latin). The exceptions to this rule are three Latin texts stemming from the Iberian Peninsula (155–157): a notarial document (155) and a sacral inscription (156), each preserved on a slate plate, and a dedicatory text preserved on a limestone slab (157).
The texts included in this survey have been excerpted on the basis of a rather simple criterion: to qualify for inclusion, epigraphic documents of practical magic written in Greek and Latin were required to present at least one verb of adjuration (ὁρκίζω and its prefixed variants in Greek, adiuro/coniuro in Latin) in a performative formula.15 Simply put, performativity in our context means that the formula is not merely transmitting information, but is intended to produce very specific yet variegated results, ranging from exercising influence on the outcome of chariot races (e.g. 64, 125) and curing the beneficiary from fever or other ailments (e.g. 8, 19) to gaining the affection of a love interest (e.g. 47, 120) and protecting the sanctity of the final resting place of family against grave robbers and/or repeated burials (virtually the entirety of the funerary inscriptions employing performative adjuration formula, 72–119; 133–154). The corpus of excerpted texts has been put together by a combination of using digital databases (see above), as well as rather meticulous analogue methods of scouring indices of epigraphical corpora, which has been necessary especially in the case of Greek material, given the wide variety of spellings of the verb of adjuration.16
While the requirement for an unambiguously attested ὁρκίζω/ὁρκῶ in Greek and adiuro/coniuro in Latin was necessary for heuristic purposes, we must point out that functional equivalents of a performative adjuration formula may occasionally be identified even in instances where no ὁρκίζω/adiuro has been used.
First, semantically related verbs, most notably the Greek ἐπικαλοῦμαι, might be used in a formula that is structurally identical with a formula introduced by ὁρκίζω/adiuro (37, 46, 56, 67); indeed, at times several semantically related verbs are grouped together to create an amplification effect in expressions such as ἐπικαλοῦμαι κὰ παρακαλῶ καὶ ἐξορκίζω ὑμᾶς (‘I call upon and I exhort and I adjure you’; 11) and ὁρκίζω σε καὶ ἐναρῶμαί σε καὶ ἐνεύχομαί σ‹σ›οι (‘I adjure you and I implore you and I pray to you’; 36). In Latin documents, we may cite the sequence adiuro te demon quicunque es et demando tibi (‘I adjure you, daemon, whoever you are, and I command you’) on a North African defixio (125)17 and coniuro vos et contestor per Patrem (‘I conjure and appeal to [you] through the Father’), preserved on a mediaeval amulets from Scandinavia (167, 171).
Second, a verb of adjuration might be omitted entirely, yet without the allocution losing its performative force. By way of example, a magical gem preserved in the Biblioteca Universitaria Valencia, dated to the 3rd to 4th cent. CE, reads τὸν θεόν σοι τὸν ὕψισ{σ}τον μή με ἀδικήσις on Side A and μέγα τὸ ὄνομα on Side B.18 The beginning of the inscription on Side A must be understood as (ὁρκίζω) τὸν θεόν σοι τὸν ὕψισ{σ}τον (‘I adjure you by the Highest God’), with the omitted verb governing the accusative that follows.19
To qualify for inclusion in the survey, the verb of adjuration also had to be attested in the text of the inscription proper, since editorial restitutions of verbs of adjuration are often highly conjectural.20 Examples with such restitutions include: (1) a fragment of a Roman terracotta lamp from Corinthus with an apotropaic formula, dated to the 6th cent. CE;21 (2) a fragment of a Christian liturgical exorcism on a silver lamella from Cyprus, invoking Jesus Christ, his apostles, and angels (possibly 4th cent. CE);22 (3) a Christian titulus from Sparta, a fragment of a marble stele dated to the 5th to 6th cent. CE, mentioning the Trinity and possibly threatening potential wrongdoers with the fiery wrath of God;23 (4) an undated Christian titulus from Lesbos, possibly invoking a ‘martyr’;24 (5) a pagan epitaph for Annia Eutychis from Thessaloniki, engraved on a sarcophagus and dated safely to the 2nd to 3rd cent. CE, apparently adjuring ‘heirs’ and ‘all the gods’ to prevent repeated burials;25 and, finally, (6) a much damaged, undated funerary inscription from Olympos (Lycia), invoking God Almighty, also to prevent future burials (and threatening a monetary sanction should someone violate the beneficiary’s wish).26
Another item not excerpted in our survey is the so-called Phalasarna tablet, a puzzling apotropaic text from Crete containing the sequence αρκομεμπομπα[..]ετωικυνε, interpreted by David Jordan as Ὁρκõμεμ ΠΟΜΠΑ[..]ΕΤΩΙ κύνε.27 To be sure, it is not impossible that the verb of adjuration (ὁρκοῦμεν) is here being used performatively, but the general sense of the adjuration is unclear and did not warrant inclusion in the corpus. The verb ὁρκῶ with performative force is attested very rarely in our survey—only 7 of the 119 Greek epigraphic documents use it (73, 74, 104, and 118 have ὁρκῶ; 96, 98, and 99 employ the prefixed ἐνορκῶ). Another complicating factor is the dating of the Phalasarna tablet. The ‘4th or perhaps early 3rd century’ BCE, proposed by Jordan,28 would predate the earliest unambiguous use of a performative adjuration formula by several centuries.29
To achieve our primary aim of producing a compact, single-volume monograph accessible to a wider audience, with all texts translated into English, we had to narrow the focus to epigraphical documents involving practical magic and performative funerary adjurations, thereby also excluding material that is closely related to the spells and curses excerpted in our survey, namely handbooks and instructions for the preparation of spells found in the (rather inappropriately named) Greek Magical Papyri (PGM), edited in two volumes by Karl Preisendanz. In keeping with our methodology, we included only those items in PGM that count as documents of practical magic,30 as opposed to written instructions on how to prepare apotropaic amulets or curse tablets.31
The dividing line between the two groups is thin, to be sure, especially considering less experienced practitioners who miscopied the instructions found in magical handbooks (e.g. 9) or failed to replace the generic placeholder δεῖνα (‘so-and-so’) with the name of the beneficiary (e.g. 28). The Greek documents are ripe with such errors, and we occasionally find them also in Latin texts—one interesting example is an amulet from Schleswig (164) where the writer apparently copied the entirety of the apotropaic spell from a formulary and forgot to include the name of the person that the amulet was supposed to protect, writing only N (possibly short for nomen and equivalent to the Greek δεῖνα).32 That being said, markers of practical use do exist and include the insertion of the proper names of the beneficiaries (and/or targets in the case of curse tablets) or traces of folding, which either compressed the size of the papyrus or thin sheet of metal to make it easier to wear in a small container (in the case of amulets) or served as part of a ritual action (in the case of defixiones, with the folding of the tablet often followed by it being pierced with a nail).
The material in the survey is organized by language (Greek, Latin) and document type (amulets, curse tablets, funerary inscriptions); within the resulting eight chapters, the entries are grouped together by location rather than date of origin. Each entry represents a single object, with several exceptions, most notably the ‘serial curses’ from Amathous (59) and Rome (62). Since several adjuration formulae may be found on a single object,33 the total number of adjuration formulae is significantly higher than the number of entries (171 in total). Throughout this monograph, individual entries found in the survey are referred to by their number in bold. The Greek and Latin texts follow their respective editions as closely as possible, notwithstanding a few minor modifications necessary to consolidate the text.34 All source texts are translated into English—whenever an English translation existed, we adopted it (sometimes with minor or major modifications) to eliminate redundancies; many texts unavailable in modern languages received fresh English renditions.
Individual chapters in the survey are organized in the following manner: The first criterion is language, with Greek (chapters 2–4) and Latin (chapters 5–8) material grouped together. Within each language category, entries are grouped according to the document type and ordered chronologically. We therefore start with Greek amulets (Chapter 2) with the earliest safely preserved performative adjuration formula on an amulet from Beirut, dated to the 1st cent. BCE or 1st cent. CE (6), and continue with Greek curse tablets (Chapter 3), Greek funerary inscriptions (Chapter 4), Latin defixiones (Chapter 5), Latin tituli sepulcrales (Chapter 6), Latin Phylacteria and varia (Chapter 7) and, finally, Latin Mediaeval Amulets (Chapter 8), which, notwithstanding a single exception (159), are from the Middle Ages. The Latin documents are always included in extenso; the Greek epigraphic material has been, on occasion, abridged due to the extraordinary length of several phylacteria and defixiones.
1.2Chronology, Geography, Language
All inscriptions excerpted in our survey are dated to the Common Era; possibly apart from an amulet against a wandering womb, a Greek-inscribed golden lamella from Beirut (6), datable to the 1st cent. BCE or 1st cent. CE and, as such, the earliest undisputable attestation of a performative adjuration formula in an epigraphic document. Another very early find, dated to the 1st cent. CE, is a Greek-inscribed wooden tablet from Sakkara (45). The earliest Latin attestations of performative adjurations are from the 2nd cent. CE, including defixiones from Carthage (123) and Hadrumetum (124). In contrast, the most recent Latin inscriptions excerpted in the survey are mediaeval amulets from Denmark, dated between the 14th and 16th cent. CE (167, 168). Greek epitaphs featuring ὁρκίζω/ὁρκῶ start appearing from the 2nd cent. CE onwards in Asia Minor, Africa, and Macedonia; the earliest Latin funerary inscriptions with performative adiuro/coniuro are attested from the 4th cent. CE in Italy and Dalmatia. Greek pagan epitaphs form a compact group of inscriptions datable consistently to the 2nd to 4th cent. CE; Latin tituli are exclusively Christian and range from the 4th to up to the 9th cent. CE. Another relatively uniform group in our survey consists of curse tablets, attested from the 1st to the 4th cent. CE, while the zenith of the use of adjuration formulae falls within the 2nd and 3rd cent. CE (35 instances in total). In the group of Greek amulets, the most numerous are documents datable to the 4th or 5th cent. CE (17 instances).
We may conclude that inscriptions featuring a performative adjuration formula are most frequently attested in the first three centuries of the Common Era: of the 90 instances, we find 42 curse tablets, 37 epitaphs, and 11 amulets. In the following three centuries, roughly coextensive with the period of Late Antiquity, the situation began to change: an increase in the number of amulets featuring such adjurations (23 instances) was accompanied by a substantial decrease in the number of curse tablets with such adjurations (only 7 instances); the number of funerary inscriptions featuring adjurations remained the same as in the previous period (31 instances). Within the transformational epoch of the 7th to 10th cent. CE, only eight items have been identified—one amulet, three epitaphs from Italy, and four Visigothic slate plates. Finally, from the 10th cent. CE onwards, we excerpt 14 Latin documents with adiuro/coniuro, esp. from Central Europe and Germany (six documents in total, 160–165) and Scandinavia (also six items, 166–171); two performative uses of adjurations from this epoch are attested from Spain (157, 158). Except for a dedicatory inscription from Celanova (157), all of the remaining documents are Christian amulets with texts inscribed on lead tablets.
The geographical distribution of inscriptions with an attested performative adjuration formula (cf. our Map, ill. 1 in 1.3.1) is remarkably wide and includes most of the regions of Asia Minor (Galatia, Cilicia, Lycia, Phrygia, Bithynia, Pontus, Sarmatia), the provinces of Syria and Iudaea, and the Greek provinces of Achaia and Macedonia; several epigraphical documents are attested from Cyprus. The largest number of texts with adjurations stems from Egypt and North Africa, 49 instances in total, comprising 13 amulets, 34 curse tablets, and 2 epitaphs. These attestations are all from the 1st to 6th cent. CE; the most recent of them are amulets preserved on papyrus (26, 29–30), while the culmination of production is to be placed in the 2nd and 3rd cent. CE (30 instances). From the region of Italy, we have 32 texts with performative adjurations (7 amulets, 3 curse tablets, and 22 funerary inscriptions), datable to the 1st to 5th cent. CE. Asia Minor supplied 31 inscriptions for our survey, comprising 5 phylacteries, a single defixio, and 25 epitaphs—most epitaphs (21 instances) form a relatively compact cluster from the 3rd cent. CE. More numerous finds are attested from mainland Greece and the Aegean islands (14 instances), Macedonia (6 instances), and Cyprus (4 instances); no less than 7 entries are from Syria, 4 from the Iberian Peninsula, and 3 each from Iudaea and Dalmatia. Regions where the use of a performative adjuration formula has been (to date) attested only rarely include the British Isles and the province of Sarmatia (one instance each). Mediaeval Latin amulets come from the regions of modern Scandinavia (6 items), Germany (5 items), and the Czech Republic (1 item). Four inscriptions, an amulet (9), two gems (33–34), and one defixio (70), are completely unprovenanced.
From the southern provinces of the Roman Empire, we have only Greek inscriptions from Egypt and both Greek and Latin inscriptions from the regions of Africa Proconsularis and Africa Byzacena. In Greek documentation, we find all three main categories—amulets (10–21), curse tablets (45–57, 63–69), and, rarely, epitaphs (114–115). Latin inscriptions from Northern Africa are exclusively defixiones (120–132). As regards Italy (21 inscriptions), Sicily (2 inscriptions), and Sardinia (1 inscription), the epigraphic documents featuring adjurations there are mostly written in Latin and are limited to epitaphs (133, 134, 137–151); Greek documents from these regions (usually from Sicily and Rome) include several amulets (24–30), two defixiones (60–61), and three epitaphs (117–119). Rare attestations of Latin inscriptions in a funerary context are from Dalmatia (152–154). Further to the West, we find only Latin epigraphic documents from the Iberian Peninsula (155–158). The unique find of an amulet against a wandering womb from West Deeping (159) presents the only attestation of a performative adjuration formulae from the territory of the British Isles as well as the only such instance of a Latin-inscribed amulet from Antiquity. Mediaeval documents with a performative use of adiuro/coniuro come from Central Europe (6 items) and Scandinavia (also 6 items).
Of note are the texts that combine Latin and Greek—in the case of (predominantly) Latin-inscribed curse tablets, the use of Greek is usually reserved for voces magicae (121–123, 127–131). In some curses from Northern Africa, Latin texts are written with the Greek alphabet (120, 124); the inverted procedure has been employed in the case of a bilingual agonistic curse from Carthage (68) where the spell text, inscribed mostly using the Greek alphabet, is introduced by a Greek text written in the Latin alphabet (horcizo se daemonion pneumn [sic] to enthade cimenon to onomati to agio). On several occasions, we find bilingual texts combining Greek with Hebrew and Aramaic, especially in the case of amulets (8, 12); one such example with an adjuration formula is also attested for Latin inscriptions, in the rather well-known epitaph for Aurelius Samohil from Catania (149).
1.3Materiality, Location, Manipulation
Performative adjuration formulae from Graeco-Roman Antiquity and the Latin Middle Ages have been preserved on a wide range of material supports. Defixiones or curse tablets were usually inscribed on thin sheets of lead, but we also find bona fide curses with adjurations on other supports, such as ostraca, papyri, and wooden tablets (cf. Section 1.3.1). Adjuration formulae in a funerary context are as variegated as the materials used to fashion steles, sarcophagi, and other means of marking the final resting place of the deceased (cf. Section 1.3.2). In the case of apotropaic texts, material supports range from precious materials (gold, silver, bronze) to lamellae made of more mundane stuff, such as copper and lead. The amulets are not, however, limited to metal supports, as spells with a protective magical nature have been also frequently inscribed on papyrus and semiprecious gemstones; phylacteria designed to prevent the onslaught of the elements are often attested on rock supports (limestone or marble; cf. Section 1.3.3). In the three sections that follow, we detail the richness of the possible material supports for epigraphic texts containing an adjuration formula, and also the typical placement of these objects and the varieties of their ritual manipulation.
1.3.1Curse Tablets
The text for virtually all defixiones excerpted in our survey was engraved into lead tablets, and all exceptions to this rule come from Egypt. An anger-restraining spell from Hermopolis (53), dated to the 3rd cent. CE, used an ostracon as the material support, and two further spells were preserved on papyrus—a homoerotic love spell from Crocodilopolis (2nd cent. CE; 49) and a love spell of considerable length and complexity from Lycopolis (5th cent. CE; 57), by which a certain Theon hoped to attract Euphemia. The latter find is extraordinary also due to the magic paraphernalia found alongside the papyrus. Two wax figurines (kolossoi), seemingly in a position of mutual embrace, were discovered wrapped by the very papyrus containing the spell, a rare instance of a love spell carefully effectuated according to the instructions found in magical formularies, closely mirroring the ‘Wondrous Spell for Binding a Lover’.35 Another unique discovery is collated under our 51 (Oxyrhynchus, 3rd cent. CE): two defixiones by the same hand on lead tablets (51.1, 51.2), found inside a clay vessel, which itself was inscribed with a spell containing an adjuration formula (51.3). One of the earliest attestations of a performative adjuration, a prayer for justice from Sakkara (45) dated to the 1st cent. CE, has been preserved on a diminutive wooden tablet.
Once a curse tablet was inscribed, it was often ritually manipulated by means of folding, rolling, or piercing with a nail.36 In the group of defixiones excerpted in our survey, 21 items feature no reported traces of manipulation; amongst those that do, 13 instances show only rolling (predominantly from Egypt and Africa), 9 tablets were only folded, and 6 further lamellae betray traces of piercing alongside folding and/or rolling—these latter come from Egypt (47, 48, 52, 56), Cyprus (59), and Corinthus in the Greek mainland (36). A particularly violent curse from Cos (37) suggests that the tablet has been put into flames to submit its target, a certain Hermias, to a fiery punishment on the basis of the similia similibus principle.37
Once the tablet was engraved and (potentially) ritually manipulated, it had to be deposited somewhere. The most typical location for a curse tablet was the resting place of one of those who died an untimely death, since their ghosts (usually called νεκυδαίμονες, ‘corpse-daemons’), once bound to the will of the practitioners, were supposed to execute their wishes. We therefore often find curse tablets deposited in tombs and also amphitheatres as the locations of deadly combat. Also not unusual was the deposition of curses into places connected with water, such as rivers, wells, springs, or baths; often, these were considered places of ritual power—the caches of defixiones found in the sanctuary of Sulis Minerva at Bath and the Fountain of Anna Perenna in Rome are typical examples of this practice. The curse tablets found in the sanctuary of Mater Magna and Isis in Mainz38 and those discovered in the sanctuary of Demeter and Kore in Corinth39 provide instances of a sacred place not connected with the element of water. On occasion, curses were also found at places that were linked to the desired effects of the practitioners, such as an amphitheatre or a circus in the case of agonistic defixiones;40 there is also ample evidence for aggressive magic to have been placed inside the dwelling of a targeted individual.41
As regards the curse tablets collated in our survey, the specific place of deposition is unknown in 21 cases (mostly in African provinces). In 19 instances, mostly in Northern Africa (15 items), but also in Cyprus (2 items) and Italy (2 items), the defixiones were found in graves and tombs, which is unsurprising, since in these cases the corpse-daemons are often explicitly mentioned as the direct addressees of the adjuration formulae; one may cite examples such as ἐξορκίζω σε, νεκυδαίμ[ων] ἄωρε, ὅστις ποτ’ οὖν εἶ (‘I adjure you, spirit of one untimely dead, whoever you are’) from Carthage (63), closely paralleled by the Latin adiuro te demon quicunque es from Hadrumetum (125).42 Also in Carthage, several agonistic curses were found on the grounds of an ancient amphitheatre (66, 67, 121–123); for two Greek curses, an aquatic context for the place of deposition may be safely assumed (42, 44), and one Greek defixio comes from the sanctuary complex of Demeter and Kore in Corinth (36).
1.3.2Epitaphs
Funerary inscriptions with a performative adjuration formula, designed almost exclusively to protect the integrity of tombs and graves, were, for the most part, cut into the rock stelae or sarcophagi. On occasion, we find less common material supports—adjurations from Melos (75) and Syros (76) were engraved on the walls; some others, such as 87, were secondarily used as building materials. Further, we find formulae attested on a terracotta plaque (100) and a brick in the form of a tabula ansata (111), both from Asia Minor. A funerary inscription from Sicily (150), dated to the 6th cent. CE, is a dipinto on a rock face, and 154 was engraved into a round mensa. In the group of inscriptions where the material support is specified in detail, marble steles are most common (27 instances in total, often coming from Italy), followed by inscriptions directly on sarcophagi (12 instances) and tituli engraved into limestone (7 instances).
In terms of geographical distribution, two pagan epitaphs are attested from Egypt: a metrical inscription dated to the 1st or 2nd cent. CE (114), preserved on a limestone stele, and an elaborated epitaph that is best placed in the 2nd or 3rd cent. CE (115). The largest number of Greek-inscribed pagan funerary tituli are datable to the 2nd or 3rd cent. CE; the inscriptions are clustered in Asia Minor (24 instances in total). A sizeable group of earlier epitaphs (starting from the 4th cent. CE), engraved for the most part in marble, are from the region of Italy (22 instances). The most recent excerpted document is that from Ourense (157), dated to the year 936, together with two other Visigothic slate plates (155–156), likely from 7th cent. CE Spain. It is worth noting that the Visigothic material just mentioned consist of documents with a dedicatory, notarial, and sacred nature.
Over one half of the excerpted texts exhibit clear Christian markers; occasionally, we also find Jewish elements and it is likely that texts like those of 87, 91, and 149 were commissioned by Jews. The original location of many epitaphs is unrecoverable due to the destruction and reuse of their material supports (this being the case for 55 texts out of 71 funerary inscriptions in total)—those found in situ were most often found in (or in the vicinity of) graveyards or churches, as is to be expected, since the performative adjuration formulae attested in a funerary context were almost always employed to safeguard the final resting place of the beneficiaries and their families against destruction, looting, or reuse.
1.3.3Amulets
The most varied group of objects containing a performative adjuration formula with the longest diachronic range is apotropaic amulets. For our purposes, an amulet may be defined as ‘a small protective device usually worn on the body to guard against unwanted supernatural influences, such as daemonic attacks and ghostly visitations, or to provide protection and healing from specific diseases and illnesses thought to have a non-medical cause.’43 Although Roy Kotansky described in this concise way the amulets of Graeco-Roman Antiquity, the same may be said of the amulets from the mediaeval period excerpted in this survey. Apotropaic objects with adjurations are usually small, often personalized items that contain the name of the wearer, who was to benefit from its curative and/or apotropaic effects. More often than not, the texts of the amulets were inscribed on thin sheets of various metals, which were further reduced in size (by either folding or rolling) and worn on the beneficiary’s body, usually in a capsula or other diminutive container.
In terms of materials used, we find a wide range of supports. As indicated above, adjurations are attested engraved on metals of various rarities, such as gold (3 instances), silver (7 instances), bronze (3 instances), and copper (1 instance). The most commonly used material in our survey is lead (16 instances). We hasten to add that apotropaic amulets on lead supports are typical of the Middle Ages; for (Late) Antiquity, we find only four instances: Greek amulets from Achaia, 5th cent. CE (1); Cyprus, 6th cent. CE (23); and Sicily, 7th cent. CE (27) and a Latin uterine amulet from West Deeping, Britain, 4th cent. CE (159).
The lead amulets collated and commented on in our corpus, the use of which spanned over one millennium (4th to 14th cent. CE), were, apart from 159 (see above), preserved in regions beyond the range of the Roman Empire during its greatest territorial expansion. About one half of the mediaeval phylacteries with adjurations come from Central Europe (160–165) and the other half from Denmark (166–171). All attested objects served to protect the wearers from illnesses, daemonic incursions, and diabolic interference. In those cases where the exact provenance is known, the amulets were frequently found either directly inside individual graves (161, 162) or in the vicinity of churches and/or graveyards (160, 163, 166). The practice of folding or rolling the thin sheets of metal on which the apotropaic texts were inscribed remained unchanged in the Middle Ages. In the case of the amulets from Salhausen (161) and Ummendorf (163), the corners of the lead tablets have been perforated to facilitate their wearing by the beneficiaries.44
Returning to the period of (Late) Antiquity and Greek-inscribed amulets, gold (in the form of a small, thin sheet) is present three times in our corpus: it carries an apotropaic spell against a wandering womb (6), found in Beirut and dated to the 1st cent. BCE to 1st cent. CE, which was folded and found inside of a tiny capsule (also made of gold); a personalized protective spell to benefit one Schybos, son of Marylleina (25), with a Sicilian provenance (5th cent. CE), featuring clear traces of folding; and, lastly, a thin sheet of gold (32), likewise originally folded, possibly from Asia Minor and best dated to the 4th to 5th cent. CE.
Silver as the material of choice for apotropaic amulets is attested more frequently than gold (7 times); all instances excerpted in the survey are Greek-inscribed and have a chronological range of the 3rd to 5th cent. CE. Two objects have been found in Egypt (10, 12), one in Syria (7), three in Italy (28–30), and a single piece in Sarmatia (31). As is the case with the amulets on gold sheets, the apotropaic spells inscribed on silver lamellae present traces of folding or rolling. By way of example, the amulet from Syria (7) was folded and inserted in a cylindrical container and the phylactery from Capua (28) was rolled and worn inside a bronze cylindrical capsule, as was another amulet from the same region (29).
Three Greek amulets employed bronze as their material support and include a phylactery from Asia/Phrygia (3), dated to the 4th cent. CE; an amulet from Lycia (5), dated to the 5th cent. CE; and a coaevous protective circular medallion from Sicily (24). Both amulets from Asia Minor were discovered in graves, while 3 was found inside a silver tubular container. Of the objects excerpted in our survey, we found only a single instance of an amulet inscribed on a copper sheet: the lamella from Iudaea (8), dated to the 4th or 5th cent. CE, likely comes from a grave and exhibits traces of rolling, possibly itself a trace of it being worn in a capsule by its beneficiary.
In addition to various metals, we also find phylacteria preserved on papyrus (9 instances) and gems (6 instances). The papyrus group is relatively compact—all objects come from Egypt and are datable to the period of ca 4th to 6th cent. CE. More often than not, amulets on papyrus contain longer and more complex apotropaic spells that feature voces magicae and other magical signs, including the omnipresent charaktêres (cf. esp. 11–13, 20). Of the papyrus amulets, two are possibly pagan (14, 19); both were intended to protect the wearers against fever. Another two amulets, both dated to the 5th cent. CE, present Jewish or Judaeo-Christian markers—such as references to the ‘God of Abraham, Isaac, and Jacob’ (11) or King Solomon (21). Three amulets are clearly Christian (13, 15, 16) and two later phylacteria (20, 22), dated to the 6th cent. CE, are ambiguous, but possibly to be ascribed to the Christian milieu as well.
Most of the papyrus amulets contain traces of spatial diminution. The one from Heracleopolis Magna (14), dated to the 3rd to 5th cent. CE, was found folded and attached to a small cylinder wrapped in a red thread, and an unprovenanced papyrus sheet with an apotropaic formula (22) was folded six times, undoubtedly to then be inserted into a container; two other papyri likewise carry traces of folding (19–20). For the most part, the beneficiary’s name is included directly on the amulet itself. The exceptions to this general tendency are a victory charm (11), an amulet miscopied from the magical handbook to ‘protect a certain man or a certain woman’ (13), and two amulets where the beneficiary is denoted in an impersonal way as the ‘wearer’ (15–16)45. The purpose of the amulets preserved on papyrus was always apotropaic, with individual spells singling out the most dangerous threats—to wit, 15 and 22 protect the wearer against the incursion of daemons, while others explicitly mention the specific ailments of a fever or migraine (13, 14, 19, 20); one amulet was supposed to guard the beneficiary against poisonous animals (16).
Six apotropaic adjurations, all written in Greek, are attested on gems (2, 9, 17–18, 33–34), which may be collectively dated to the 2nd to 3rd cent. CE. As is often the case with magical semiprecious gemstones, there is little indication of their exact provenance. In half of the cases, we possess no indication at all (9, 33–34), while 2 may possibly come from Asia Minor and 17–18 from Egypt. Due to their minuscule size, gems usually contain only short texts and none of the six objects excerpted in our survey is personalized. Of particular interest is a red carnelian (9), once owned by W. Froehner and today preserved by the Bibliothèque nationale de France; the text is unusually long and complex, and the spell copied therein is expressis verbis defined as an ‘adjuration’: Ὁρκισμὸς οὗτός ἐστ(ι) Σαβαὼθ Ἀδωναί, τοῦ μὴ ἐγγίσαι ὅτι Κυρίου θεοῦ Ἰσραήλ (‘This spell/adjuration involves Sabaôth Adônai. Do not come near, because [the owner] belongs to the Lord, God of Israel’).
The apotropaic amulets found in situ point to the common practice of burying the deceased together with his/her amulet, especially where the protective device has been personalized via inclusion of the beneficiary’s name—one may hypothesize that an amulet that has been personalized, with its curative and protective magical power focused exclusively on one particular individual, would turn out to be ineffective when worn by anybody else (cf., for instance, 7, 12, 31). As indicated above, the practice of burying the dead with their amuletic paraphernalia continued also in the Middle Ages, see 161, 162.
The amulets were not designed to protect only individuals, however. In our survey, we excerpt also three phylacteries that were aimed to protect the livelihood of their beneficiaries, usually fields and vineyards, against the onslaught of the elements. These include two Greek amulets—a large marble plaque from Lydia (4), dated to the 5th cent. CE, and a polygonal limestone from Sicily (26), dated to the 5th cent. CE— and one Latin example, an amulet for land protection on a Visigothic slate plate from the Iberian Peninsula (158), dated to the 9th to 10th cent. CE. Phylacteries designed to protect agriculture are often larger than amulets fashioned to be worn on the person of the beneficiary (for instance, the amulet to prevent hailstorms from Lydia measures 68 × 33 cm), and they were most likely placed inside buildings or buried in the very fields they were supposed to protect. A miscopied instruction on the Lydian phylactery provides additional detail as to the location of these protective devices, since it indicates it was to be ‘placed in the three corners of the vineyard’ (τιθὶς ἰς τρῖς γονίας τοῦ ἀνπελõνος).
1.4Structure, Function, Syntax
‘I conjure you by that which you profess, / Howe’er you come to know it, answer me.’46 It is with these words that Macbeth, consumed by anxiety and terrified at the prospect of losing his ill-begotten throne, turns to the witches at the beginning of Act 4 in Shakespeare’s eponymous play. The formula that the doomed king employs starts with a verb of adjuration (‘I conjure’), used performatively in the first-person singular, and continues with a direct addressee, meaning the entity or entities that are adjured, here present in the form of the pronoun ‘you’, clearly denoting the Weird Sisters.47
The formula is then expanded by a mediation syntagma in the guise of the prepositional construction ‘by that which you profess’. The meaning of this sequence in Shakespearean scholarship is often connected to the witches’ superhuman epistemic abilities and taken to mean something like ‘I conjure you by that which you claim to know, whatever the source of your knowledge is’,48 yet it is more likely that the syntagma designates an entity or entities to which the witches are beholden, be it Hecate or the apparitions. The meaning, then, would be ‘I conjure you by the power of the entity which you honour as your superior’. After all, the Weird Sisters themselves ask Macbeth whether he would prefer to hear the answers ‘from our mouths, / Or from our masters’.49 Following the prepositional clause, we find the desired effect of the adjuration—a command to the witches to unveil the future (‘answer me’). When the king’s demand to obtain one final assurance (‘shall Banquo’s issue ever / Reign in this kingdom?’) is met with resistance (‘Seek to know no more’), he adds a sanction: ‘I will be satisfied. Deny me this, / And an eternal curse fall on you’.50
Despite Shakespeare’s inspiration being primarily literary51 and not epigraphical, it is remarkable that we find in a fully formed performative adjuration formula with all its constituent parts in a theatre play written and staged more than a millennium after the flowering of the Mediterranean magical κοινή in which these illocutions bloomed in the hundreds. In what follows, we offer a brief outline of the main building blocks of adjuration formulae, which are then consistently used in the commentary section of the survey. It is hoped that the nomenclature developed and systematically employed here will help establish a common terminological ground that will facilitate the discussion of performative utterances in Graeco-Roman magic and beyond.
1.4.1Verb of Adjuration
We have already established that the verb of adjuration served as the principal criterion that all texts excerpted in this survey had to satisfy (cf. Section 1.1). The required element of performativity dictates that the verb of adjuration is virtually always attested in the first-person singular (most common) or plural (rather rare)—the only two exceptions to this rule in Greek documents are the forms ὥρκωσαν (3rd pers. pl.), attested in a metrical inscription from Tanagra (74), dated to the 5th cent. CE, in which the aorist might have a gnomic, timeless value, with the sense of ‘we have adjured’ (at some point in the past and the adjuration is now valid indefinitely),52 and ἐξορκίζει (3rd pers. sg.) in an epitaph from Elaioussa Sebaste (105), dated to the 2nd or 3rd cent. CE, which forms part of a highly unusual adjuration that lacks both direct addressee and mediation syntagma. In the texts of later Latin epitaphs, the use of the verb coniuro in the 3rd pers. sg. is attested thrice (135, 142, 157); we also find the participle coniurantes (139) and the perfect coiiurabit (= coniuravit), likely with gnomic value (154). The employment of the 3rd pers. sg. in a performative adjuration formula (instead of the prototypical 1st pers. sg.) is due either to an error on the part of the scribe or (much more likely) a self-reference in the third person that builds the adjuration on preexisting syntactic scaffolding without losing its performative force.
Since ‘serial’ adjurations occasionally employ the same verb of adjuration dozens of times,53 a better metric than the sum total of all attestations of any particular verbal form is the number of instances in which at least one particular form is used in any given document type. In the texts written in Greek, comprising 119 entries in our survey, the most commonly used verb of adjuration is ὁρκίζω (used 50 times), followed by ἐξορκίζω (used 39 times) and ἐνορκίζω/ἐνορκίζομαι (used 21 times). It may be concluded that the unprefixed ὁρκίζω, together with prefixed forms with ἐξ- and ἐν-, dominates the field, since other verbal forms are much less frequent: compounds with ἐπι- are attested five times (7, 18, 33, 35, 103) and compounds with κατα- twice (107, 111); we also find two forms with a double prefix: προσεξορκίζω (65) and ἐπεξορκίζομεν (94). Adjurations employing a verbal form with the same root but lacking the productive -ίζω suffix are employed seven times in total—we find the unprefixed ὁρκῶ (attested four times, 73, 74, 104, 118) as well as a prefixed form, ἐνορκῶ (attested three times, 96, 98, 99).
Of note is the fact that amulets and curse tablets only rarely use verbs other than ὁρκίζω and ἐξορκίζω—we find only two instances of compounds with ἐν- and four instances of compounds with ἐπι- in the group of 70 excerpted phylacteria and defixiones, notwithstanding a double-prefixed προσεξορκίζω (65). This is in striking contrast with the formulae in a funerary context, where compounds with the prefix ἐν- are attested most often: out of 49 Greek tituli excerpted in the survey, almost half of them have ἐνορκίζω/ἐνορκίζομαι (attested 19 times) or ἐνορκῶ (attested 4 times), compared with 14 instances of ὁρκίζω and 6 instances of ἐξορκίζω. The forms ὁρκῶ/ἐνορκῶ are attested only in the funerary context, and ἐνορκῶ in particular seems to be geographically conscribed to the region of Galatia in Asia Minor (96, 98, 99).
It would be tempting to try to single out specific semantic content for particular forms (such as ‘I cast out, I drive out’ for ἐξορκίζω, as opposed to ‘I invoke, I call upon’ for ὁρκίζω), but the epigraphic material that has been collated in this survey speaks strongly against any such assumption. Notwithstanding a clear preference for forms with the prefix ἐν- (ἐνορκίζω, ἐνορκίζομαι, ἐνορκῶ) in a funerary context, any verb of adjuration may be used to express a wide spectrum of performative functions, ranging from a humble petition to a straightforward command that threatens punishment if the desired effect is not brought about (cf. Section 1.4.5).
For instance, the form ἐξορκίζω, far from being a staple for casting out daemons, is routinely used to call them forth (as is often the case in defixiones; cf. 47, 49–52, 55, 57, 60–67, 69) and may even be employed in an invitation to dance—the formula ἐγώ, Kλευπάτρα, ἐξορκίζω σε τὸν Kάβειρον, ἀναγνοὺς ὀρχῆσε (‘I, Cleopatra, adjure you by Kabeiros, read and dance!’), inscribed on a marble stele in Thessalonica sometime in the 2nd cent. CE, has precious little to do with banishing illness-causing daemons or their being coopted by practitioners of magic to fulfil desires; rather, it is a playful reminder to enjoy the good things in life while they last, kindly provided by Cleopatra, who may no longer delight in such simple pleasures. Even compounds that are rarely attested, such as verbal forms with ἐπι- (five instances), are employed across a variety of different material supports and spanning all three major document types, including a silver apotropaic lamella (7), two magical gems (18, 33), a defixio written on a lead sheet (35), and a funerary inscription (103).
In the documents written in Latin, the verb adiuro is attested in ca two thirds of cases; the remaining instances employ coniuro. Their performative force is at times reinforced by use of semantically neighbouring verbs—we may cite sequences such as adiuro et demando (127–131; frequently used in serial defixiones from Northern Africa) and adiuro/coniuro et contestor in the texts of mediaeval amulets (167, 171). The verb adiuro is used in all curse tablets excerpted in this survey; coniuro is more typically found in earlier documents, especially Christian epitaphs designed to protect tombs.54 In the case of amulets, adiuro again takes precedence over coniuro, but some texts combine the two verbs (such as the phylactery for a boy named Tado from Halberstadt, our 162). The texts excerpted in our corpus show that there is little if any semantic difference between adiuro on the one side and coniuro on the other. Each is typically followed by a prepositional mediation syntagma introduced with per (cf. Section 1.4.3) and a subordinate clause with ut/ne/non; at times, the desired effect is expressed asyndetically via the subjunctive or imperative (122–123, 126; cf. Section 1.4.4).
1.4.2Direct Addressee
By direct addressee, we understand the entity or entities the petitioner reaches out to in order to fulfil his/her wishes; syntactically, it is (almost always) the direct object of the verb of adjuration, namely the accusative that follows immediately after ὁρκίζω/adiuro and the like. In the Greek material, we occasionally find a construction with the dative, such as ἐνορκίζομαίσοι τὸν Θεὸν τὸ̣ν κτίσαντα τὴν γ̣ῆ̣ν καὶ τοὺς οὐρανούς (‘I adjure you by God who founded the earth and the heavens’; 3) and ἐνορκίσζομαί σοι τὸν Σεβάσστιον ὅρκ{ι}ον (‘I adjure you by the oath of the Emperor’; 79). Whenever the dative is used, the governing verb features the ἐν- prefix, which explains the attraction of a different case than the accusative; alternatively, one may interpret the dative as dativus commodi/incommodi, or even as dativus ethicus.
The direct addressee is present in all but a handful of performative adjuration formulae. In about half of these cases, the absence of the category of direct addressee is due to fragmentary nature of the inscription, lacunae, and/or illegible text (38, 71, 77, 81, 90, 102); we may also include in this group a magical gem invoking Arbath Iao and voces magicae (33) on which we find the most concise adjuration formulae imaginable, limited to an isolated ἐπορκίζω
