Navigation – Plan du site
Autres articles

An Umayyad period magical amulet from a domestic context in Jerash, Jordan

John Møller Larsen, Achim Lichtenberger, Rubina Raja et Richard L. Gordon
p. 369-386

Résumés

Durant la campagne 2014 du projet dano-allemand « Quartier nord-ouest de Jérash », une petite capsule à lamelle de métal fut trouvée dans une maison détruite par un séisme au milieu du viiie s. La capsule, qui conservait des traces de fibres textiles, était bien conservée et contenait une bande de feuille d’argent (lamelle) étroitement enroulée qui, extraite, se révéla porter des lettres d’allure sémitique. Comme elle ne pouvait pas être déroulée manuellement, elle fut soumise à une tomographie par ordinateur et « déroulée » numériquement grâce à un logiciel spécial. Cet article présente la lamelle et sa capsule ainsi que le contexte archéologique, une maison privée scellée par un séisme au milieu du viiie s. sur la terrasse est du Quartier nord-ouest de Jérash. La lamelle, qui contient aussi quelques signes (des signes magiques), est interprétée comme une amulette s’inscrivant dans la tradition sémitique aussi bien que gréco-romaine mais qui, dans ce cas, appartient sûrement à la période omeyyade.

Haut de page

Texte intégral

Introduction

  • 1 For introductory literature on Gerasa and further references see: Kraeling 1938; Zayadine 1986; Lic (...)
  • 2 On this earthquake see Tsafrir & Foerster 1992.
  • 3 Walmsley 2008 and Walmsley 2011 as well as Walmsley et al. 2008 for considerations on the Abbasid m (...)
  • 4 Lichtenberger & Raja 2016 for a presentation of the Mamluk evidence from the Northwest Quarter in J (...)

1Jerash, ancient Gerasa, a multi-period site in northern Jordan, has been the focus of much research over the last century (fig. 1). From the 1920s, when large-scale excavations were carried out by Yale University, consisting of teams also involving British archaeologists, until today the site has undergone intensive study of its urban history in several periods. 1 The development of the site in more recent times when a modern settlement began to expand on the eastern side of the Wadi Jerash, the ancient Chrysorrhoas (the Golden River), has meant that archaeological exploration is now focused almost exclusively on the western side of the wadi, where the most monumental complexes dating to the Roman period are located. The site flourished from the 1st cent. ce into the early Islamic period. Although the earthquake of 749 ce devastated large parts of the city, 2 some evidence has recently been found for a degree of continuation into the Abbasid period, 3 but substantial archaeological evidence only begins to occur again in the Ayyubid-Mamluk period, when several housing complexes were constructed in different parts of the city. 4

Figure 1.

Figure 1.

Map of Jerash with Northwest Quarter marked (after Lepaon 2011)

  • 5 Clarke & Bowsher 1986 for a short contribution on soundings in the Northwest Quarter, which yielded (...)
  • 6 Lichtenberger & Raja 2012 and Kalaitzoglou, Kniess, Lichtenberger et al. 2012 for publications on t (...)
  • 7 Lichtenberger & Raja 2015a.

2Since 2011 the Danish-German Northwest Quarter Project has been exploring the highest area within the city walls (fig. 2), which until then had remained largely unexplored. 5 This area of approx. four hectares is located behind the Roman period Artemision on a hill overlooking the city. 6 The area is densely covered with structures, which according to the examinations and excavations undertaken until now point to a surge in activity in the early third cent. ce and continuing into Late Antiquity and the Umayyad period. 7 So far no traces of an Abbasid period settlement in the Northwest Quarter have been found, and it is only in the Ayyubid-Mamluk period that substantial archaeological evidence in the form of buildings and pottery re-appears.

Figure 2.

Figure 2.

Map of the Northwest Quarter in Jerash with the location of all trenches excavated by the Danish-German Northwest Quarter Project in the period 2011-2015. Trench K is the trench in which the lamella was found.

© The Danish-German Northwest Quarter Project

The East Terrace in the Northwest Quarter

3Since 2014 the Danish-German Northwest Quarter Project has been exploring the so-called East Terrace, which is located on the easternmost edge of the hill prominently overlooking the Artemision.

  • 8 Lichtenberger & Raja 2016.
  • 9 For a more detailed description of the building see Lichtenberger & Raja et al. 2016 as well as Kal (...)
  • 10 Lichtenberger & Raja 2015b.

4During the 2014 campaign parts of a private house (trench K) on the East Terrace of the Northwest Quarter in Jerash were excavated (fig. 3). The structures on the East Terrace were heavily destroyed by the earthquake and the area is densely covered with rubble, which makes excavation of the area complicated. However, this also means that the area was sealed and that the archaeological contexts encountered during the excavation have probably been left undisturbed since the mid-8th cent. ce, thus presenting unique situations rarely encountered on sites with a long and continuous settlement history. 8 C14 dating of a number of objects from the complex excavated in 2014 showed consistent mid-8th cent. ce dates. 9 These were supported by the find of a coin hoard (located in evidence 23, fig. 3), which had been wrapped in textile or lying in a purse, in the complex. Although the latest coins from the hoard belong to the reign of ʿAbd al-Malik (685–705 ce / ah 65–86) the last coins in the associated destruction fill of the house can be dated to before 750 ce (two post-reform Umayyad coins). 10 Furthermore the latest ceramics from the complex also consistently show 8th cent. ce shapes and forms.

Figure 3.

Figure 3.

Orthophoto of trench K

© The Danish-German Northwest Quarter Project

  • 11 The lamella and its case have the find number: J14_Kh_3_498a_M113.

5The complex excavated was a private house of two storeys. On the ground floor in the westernmost part of the trench a kitchen with two hearths was located, as well as a number of other kitchen-installations and utensils. In the eastern part of the trench a forecourt with stairs leading to an upper floor was located, and a stone slab covered water-channel cut in bedrock was situated on the easternmost fringe of the trench. The upper storey had fallen onto the ground floor during the earthquake of 749 ce. It was in the find-rich debris stemming from the upper floor that the encased lamella was found. 11

The lamella and its case

  • 12 The technical aspects of this process and the analysis of the metals are described in detail by Bar (...)

6The find consisted of two parts: a container and a lamella which was placed inside the container (fig. 4-5). During conservation the metal strip lamella was removed from the container. However, the lamella was too fragile to be unfolded manually without causing extensive damage to the metal, so it was decided to submit the lamella to computed tomography followed by an attempt to ‘unroll’ the object digitally. 12

Figure 4.

Figure 4.

The container while the lamella was still inside

© The Danish-German Northwest Quarter Project

Figure 5.

Figure 5.

The container and the lamella separated after conservation

© The Danish-German Northwest Quarter Project

  • 13 Cf. for such containers Schienerl 1984.

7The capsule is 50 mm wide with a diameter of 17 mm; the lead itself is 1 mm thick. 13 No smithing lines were visible, but they may have been obscured by corrosion and mineralization on the exposed surfaces. The capsule was somewhat deformed, with a fissure running down its length and several smaller fractures, as well as holes where fragments of the metal had broken off. There were clear traces of imprints from textile fibres indicating that the container had been wrapped in cloth (fig. 6).

Figure 6.

Figure 6.

Close-up of the container showing the traces of textile fibres

© The Danish-German Northwest Quarter Project

8The rolled silver lamella (ca 0.15 mm thick and 42 mm long [fig. 5, 7]; diam. 8 mm) was extremely finely worked. The metal sheet has a calculated length of 9.5 cm. It was corroded, mineralized on the surface and small fragments from the edges were missing. When the lamella was removed from the container, it was clear that it was inscribed with letters of a Semitic character. The letters on the lamella were incised with a fine stylus with a rounded tip. The letter size can be exemplified by the initial N-shaped letter in line 6, which is approx. 3 x 5 mm (fig. 10). The lamella was folded and rolled with the front side of the text to the inside.

Figure 7.

Figure 7.

The silver lamella

© The Danish-German Northwest Quarter Project

9The container and the lamella were subjected to metallurgical analysis, which showed that the container was made of lead, possibly with some bronze decoration. The lamella on the other hand consists of almost pure silver mixed with a trace of gold.

  • 14 See Seales & Delattre 2013 as well as Mocella, Brun, Ferrero & Delattre 2015.

10The technique applied in the current case in order to digitally unfold the lamella was a combination of computed tomography and general-purpose industrial imaging software for voxel analysis (fig. 8-9). Such techniques have in recent years been applied with good results to ancient carbonized texts from Herculaneum in the bay of Naples. 14 However, the difference between the text-carriers in the two cases (papyri versus metal) as well as the use of different writing techniques (ink versus incision) raises different technical problems, and other solutions were needed for our text.

Figure 8.

Figure 8.

A slice of the lamella showing how complexly it was rolled before being placed in the container

© The Danish-German Northwest Quarter Project

Figure 9.

Figure 9.

Illustration showing the location of the slice depicted in fig. 8

© The Danish-German Northwest Quarter Project

11The results of the digital unfolding were satisfying and successful, but it was a time-consuming process involving much manual processing. The lamella yielded 17 lines of text (fig. 10-11). However, the interpretation of the text remains unclear. Nonetheless the find remains extremely significant, since it gives insight into the continued practice of producing and using protective amulets, which was an important tradition in the Semitic as well as the Greco-Roman cultural spheres.

Figure 10.

Figure 10.

Lines 1-8 from the silver lamella. Shown are the back sides, mirrored. “a” and “b” represent upper and lower parts of a line.

© The Danish-German Northwest Quarter Project

Figure 11.

Figure 11.

Lines 9-17 from the silver lamella. Shown are the back sides, mirrored

© The Danish-German Northwest Quarter Project

Philological analysis

12The letters employed on the metal sheet are not readily identifiable. It is immediately clear that two different scripts were employed. In line 1 the seven letters to the left stand out, since none is ligatured, and their appearance is different from most of the remaining text. In other lines there are a few further individual letters of a similar type, which we return to later.

  • 15 Anyhow, it is not possible to identify the lamella script with Mandaic, personal communication from (...)

13The letters to the right in line 1 and most of the writing in lines 2-16 and perhaps also line 17 clearly belong to a cursive, ligatured script. Of course, the archaeological dating of the destruction of the house in which the lamella was discovered (mid-8th cent. ce) only provides a terminus ante quem for the production of the object, but since the house was a private dwelling, it may not be unreasonable to suggest a maximum age of one hundred years for the object, which largely places it in the Umayyad period. In theory the lamella could have been brought to Jerash from a different linguistic and cultural context. However, it is safe to interpret the lamella as an amulet, and once the lamella was placed in its metal container, it had no aesthetic value. Its value was personal, and it is difficult to see that it would have been ‘imported’ to Jerash unless it was brought there by its owner. Accordingly, if we assume that it was produced locally at some point during the Umayyad period, the number of possible identifications is greatly reduced. For instance, we can leave out the Nabataean and Palmyrene scripts, which had gone out of use centuries earlier. Mandaic script, which is cursive, can be found in this period, for example on incantation bowls and metal amulets, but they are found in Mesopotamia, so that we need not entertain this possibility here. 15 Outside the Semitic script families the most obvious candidate would be cursive Greek, but this possibility can be dismissed when we look at the overall ductus of the script. The main forces driving the development of cursive scripts are ease and speed of writing, and the many oblique or curved strokes (e.g. in line 6) would only be ‘natural’ if written downwards to the left in a script running from right to left.

  • 16 Gzella 2015, p. 281-329 provides a useful overview of Aramaic in this region in Late Antiquity.
  • 17 An overview of the locations of the epigraphic material is found in Desreumaux 1998a, p. 10-11. Dis (...)
  • 18 Contrary to a widespread earlier view in research, the more cursive serto script was already presen (...)
  • 19 A chart of the Christian Palestinian script is found in Desreumaux 1998b, p. 513.

14This leads us first to the Aramaic dialects written in Semitic scripts. Samaritan Aramaic (an unlikely candidate in the first place) was written in a developed form of the (non-Aramaic) ‘Paleo-Hebrew’ script; it was not ligatured and can thus be disregarded. The two other major Aramaic dialects in use in the wider region of Palestine in this period were written in Aramaic scripts. 16 Jewish Palestinian Aramaic was written in the Jewish (‘square Hebrew’) script, but it had not developed any systematic ligaturing either. Christian Palestinian Aramaic (Melkite Aramaic), however, used a script which is clearly derived from the Syriac Estrangela script but developed its own ductus and certain peculiarities. Epigraphically this script is attested at several locations including Jerash and Khirbet el-Samra (about 30 km west of Jerash). 17 This is a ligatured script, and for the sake of argument we might consider both this western, Christian Palestinian (Melkite) script and the eastern, Syriac forms traditionally termed Estrangela, Serto and Nestorian, 18 which are all closely related and ligatured scripts. However, it is only possible to identify very few similarities between letter forms in these scripts and that of the lamella under scrutiny. In the Christian Palestinian script the letter ḥēṯ often has the shape of a Latin N; such a form is found in lines 5 and 6 of the lamella. 19 In the Estrangela, Serto and Nestorian scripts the letter qōf has the form of a loop resting on the base line; two such signs are found in line 16. A major difference between the Estrangela and Christian Palestinian scripts on the one hand and Serto and Nestorian on the other is the letter ālaf, which has very simplified forms in the latter scripts. Nothing even remotely similar to the Estrangela and Christian Palestinian forms of this very common letter is found in the text, but the Serto form, which can be as simple as a straight vertical stroke, is seen in several lines. The results, then, are very meagre, and it may be just as pertinent to look at some traits in the scroll script that decidedly point away from an identification with Syriac.

15A recurrent feature in the lamella script is letters ending in (or consisting of) a more or less pronounced curve downwards to the left. This trait is rare in the Syriac scripts where it is really only characteristic of the letter ṣāḏē. However, in another member of the Aramaic script family, Arabic, this trait is very frequent, occuring in (some forms of) the letters rāʾ/zāy, sīn/šīn, ṣād/ḍād, qāf, lām, nūn, wāw and yāʾ. We should try to pursue this possibility further.

  • 20 For examples of amulets without diacritics, see for instance Savage-Smith 1997a, no. 58 (p. 132), 7 (...)

16The script is unvocalized and without diacritical points; the first trait is quite common, while the second is not uncommon in amulets. 20 We should begin by looking at those passages that seem to be the most likely candidates for an Arabic interpretation. Line 6 begins with an N-shaped letter (fig. 10), possibly a magical sign as in lines 1 and 5 (these will be discussed separately below). Following this we may read ٮو or لو. If we read ٮو, the absence of diacritics gives us five possible interpretations: بو / تو / ثو / نو /يو ; لو on the other hand could be the conditional particle law (“if”; it can also be used optatively: “if only”, “would”). However, in line 17 لو are the last letters in the text, following the ‘trefoil’ shape, and it would of course be odd to end a text with such a particle, unless it was used as an interjection of sorts. Alternatively the two letters in line 17 could be seen as a separate magical sign similar to the one in line 1. It is noteworthy that this combination is also found in lines 8-11 and perhaps lines 5 and 7. In line 6 the top of the following letters go slightly around the bend in the sheet. When this part is viewed from a different angle in the computer program, it becomes clear that it is possible to read this either as an N-shaped sign with a long and curved left leg (and thus not as Arabic) or as ىر, where the first letter would have the same five possibilities as above, and the second letter could be read as rāʾ or zāy. The next letters could be read لىر (or لىن ?), but without diacritics the second letter has the same five possibilities as above, while the third letter again can be either rāʾ or zāy (A similar combination may be seen in line 16.) A possible reading could be li-yara “and he shall see” with the so-called lām al-amr and jussive form of raʾā. The following letter can be either rāʾ or zāy. Next, we could read an alif, but the following sign is more difficult. It could possibly be read as a lām-alif ligature, but then we cannot explain why the following rāʾ or zāy has a ligature. Finally, the sign to the left is difficult to interpret as Arabic, unless it is a poorly written lām-alif ligature. In line 10 the initial letters may be read as لن. If so, the following letter rāʾ or zāy merely touches the nūn and does not form a proper ligature. Following this we could read an alif.

  • 21 Canaan 1937, p. 81-83.

17The next two letters are also difficult. The first letter could be a lām in its isolated form; it does not form a ligature to the second letter which could be rāʾ or zāy if it were not placed above the base line. The next could again be read as an N or as ىر. Following this we have the recurrent ٮو (or لو). Finally, to the left we probably have a magical sign. We should note that the initial five or six letters in line 11 are rather similar to the beginning of line 10. In line 14 the initial letter resembles a dāl, but since it is ligatured to the following, it could perhaps be seen as a ḥāʾ. A possible reading could then be حلالرىل (with the usual diacritical alternatives). If so, could -īl be the simplified theophoric element ending an angelic name, just as Ǧabrāʾīl is interchangeable with Ǧibrīl? Angelic names found in later Muslim amulets are by no means limited to those otherwise known from the Qurʾan and tradition but could, for instance, be formed according to certain numerical rules. 21

  • 22 Kalus 1981, no. 2.10 (p. 98 and pl. 16).

18This attempt to identify meaningful text(ual elements) has not been very promising in terms of decipherment. Moreover, much of the remaining text consists of more or less vertical strokes ligatured to each other. This could mean that the text employs proper Arabic letters but not in order to convey a meaning in a natural language. If so, we could speak of a pseudo-text or a ‘magical’ text consciously employing incomprehensible syllables, although we cannot exclude the possibility of a private language known only to the practitioner who prepared the text. The use of unintelligible ‘words’ and ‘names’ is quite frequent in magical texts. In line 1 of an amulet collected in Iran, for instance, we find the following: كهيعصسىلطىحكڡمىملمڡحى, beginning with the five mysterious letters of Sura 19:1 (kāf, hāʾ, yāʾ, ʿayn, ṣād) and continuing with 16 further and incomprehensible letters. Line 2 continues in a similar vein: كهىحىعىسطىحلد لهحعىلمسلمحى, as do lines 3-4. 22

  • 23 Woltering 2015.
  • 24 Aanavi 1968, p. 353-358.
  • 25 Savage-Smith 1997a notes that “Pseudo-Kufic writing was (…) frequently employed for talismanic deco (...)
  • 26 Segal 2000, no. 121Ps-142Ps (p. 150 and Pl. 138-159); Müller-Kessler 2005, p. 5-6, 244-257.

19However, in some lines we also find forms that are decidedly not Arabic. In line 12, for instance, there is a letter consisting of an upper horizontal stroke continuing downwards at its left end; and towards the end of the same line another letter consists of a vertical stroke with a hook to the left at the top. Considering the length of the text, it is also notable that we certainly find no clear instances of the letters ṣād/ḍād/ṭāʾ/ẓāʾ/ʿayn/ġayn/kāf/mīm and hāʾ. The presence of ǧīm/ḥāʾ/ẖāʾ and sīn/šīn is doubtful as well. In view of these facts, the question arises whether the script in fact is pseudo-Arabic, i.e. mimicking parts of the graphical repertoire of Arabic script but not conveying a meaningful text, and (in this case) occasionally employing letter forms that cannot even be interpreted as Arabic. Research on pseudo-Arabic script has mostly dealt with specimens from Europe and Byzantium, 23 but the phenomenon is also widely found in the Muslim world. 24 Pseudo-script is also a well-known phenomenon in magical texts; for instance, it is found on Arabic amulets, 25 and we have numerous examples in the Aramaic incantation bowls. 26

  • 27 Kotansky 1994, no. 1 (p. 1-2), 3 (Latin, p. 13-15), 12 (p. 54-57), 14 (p. 72), 18 (p. 81-88), 41 (p (...)
  • 28 Naveh & Shaked 1987, no. 2 (p. 44-49 and pl. 2), 4 (p. 54-60 and pl. 4), 5 (p. 60-63), 8 (p. 78-81 (...)
  • 29 Naveh & Shaked 1987, bowl 1 (p. 124-132 and pl. 14-15); Naveh & Shaked 1993, bowl 26 (p. 139-142 an (...)
  • 30 Gordon 2011. A recent study (Coulon 2010) emphasizes the importance of the Greek-Byzantine traditio (...)
  • 31 Bohak 2008, p. 270-274; Winkler 1930, p. 150-167.

20As mentioned above, the text also contains a number of signs which clearly are neither Arabic nor imitations of Arabic letters. Most of these occur in line 1, i.e. the seven signs to the left. A few further signs are found at the beginning of lines 2, 5, and 6, and at the end of line 10. Line 17 contains a conspicuous trefoil-like shape. It has not been possible to assign any of these signs to a known alphabet. They are certainly neither Greek nor Latin. Since we can take it for granted that the artefact is an amulet, and since the main text seems to be in pseudo-script (characteristic of many magical texts), it is reasonable to view them as magical signs (charaktêres). The use of magical signs was widespread in Antiquity, and they are found in texts written on a multitude of media. In Greek, for instance, we have many examples from papyri, from metal amulets, 27 and from engraved gems. We have examples on Aramaic amulet texts incised on metal sheets, 28 and on occasion, magical signs are also found on Aramaic incantation bowls. 29 Frequently, such signs are of the ‘ring-letter’ type (Brillenbuchstaben), signs characterized by the addition of small rings or loops at the end of lines, but they can have numerous different forms, and the signs here belong to the simpler, unadorned forms. 30 The use of magical signs in Antiquity greatly influenced the Jewish, Christian and Muslim magical traditions, 31 but there are no obvious clues to the religious affiliation of the owner of this amulet (unless the ‘trefoil’ in line 17 could be seen as a Christian symbol).

  • 32 Walmsley et al. 2008, p. 109; see also p. 119-122, p. 124-126.
  • 33 Gawlikowski 1995 et 2004, p. 470.

21However, while the amulet may be indecipherable, it still provides important information on the cultural milieu in which it was produced. Through the Umayyad period the use of Arabic became widespread, and earlier finds have also documented Arabic texts in the early Islamic period in Jerash. For instance, two marble slabs discovered in 2007 with Arabic texts written with a charcoal pen have been interpreted as merchants’ records, testifying to “the sophistication of commerce in the eighth cent., but also the common use of Arabic (…) by this time in the market place”. 32 The use of Arabic by both Christians and Muslims has been documented on the so-called Jerash terracotta lamps which for a brief period (723–750 ce according to the dated specimens) frequently show the name of the potter inscribed in Arabic. 33 The apparent use of a pseudo-version of the Arabic script now documented shows that it was deemed suitable for magical purposes as well, but as mentioned above, we cannot identify the religious affiliation of the amulet’s owner.

  • 34 Savage-Smith 1997b, p. 60.
  • 35 See the study on such modern amulets from Jerash: Mershen 1983.

22The use of pseudo-script and letter-like magical signs is important evidence for the value and power written texts had for people in the ancient world, but it remains a question how the use of pseudo-script arose. It has been suggested that it came about “as amulet-makers copied exemplars that they did not understand. In some cases the maker, as well the person purchasing the item, may have been virtually illiterate and unable to understand the original inscription or recognize that the copy was meaningless”. 34 Therefore even if in the case of the Jerash lamella we cannot reconstruct a text from the (pseudo-)script, it is an important contribution to the reconstruction of mentalities and daily life in the ancient world. The application of the new technique on other texts written on friable or brittle rolled or folded metal sheets that cannot be opened without fear of destroying the text (such as poorly-preserved curse-tablets on lead) will provide us with further glimpses into the mentalities of people in the ancient and medieval worlds. The use of such inscribed metal amulets in the eastern Mediterranean area can be traced continuously up to modern times, also in Jerash, though nowadays they are mainly written on paper. 35

Conclusion

23The successful digital unrolling of the tightly-folded silver lamella is an important advance in technique, which may in the future also be applied to similar objects. Although the letters are clearly visible, a meaningful text could not be established, but the use of pseudo-script and charaktêres is not unusual on amulets. It is clear that once it had been inscribed, folded, rolled and placed in the container it was intended to protect the wearer from certain kinds of harm. Such amulets were not meant to be re-opened, so that the wearer was and remained ignorant of the precise text. It is plausible to suppose that the bestowal of the finished amulet was accompanied by some kind of ritual (of which we have no evidence). Since it was found in a sealed archaeological context, it may be assumed that it either was part of the personal belongings of (one of) the residents or was placed somewhere in the house for general protection.

  • 36 See the introductions by Foss 2008 and Album & Goodwin 2002.

24The lamella attests to the power of writing in Late Antiquity and the Early Islamic period. It is clear that the practitioner who prepared the amulet and possibly also the purchaser had some idea of the general appearance of Arabic script but apparently no reading competence in Arabic. It is thus likely that it was produced by a person with a different Semitic or with a Greco-Roman background. An apparently meaningless text could be said to be in contrast to the valuable material of the lamella, which was silver with a trace of gold. This, however, can be compared to other contemporary phenomena such as Arab-Byzantine coinage, which also sometimes carries imitations of Greek letters. 36 It is clear that script, medium (the coins) and material in such cases belonged together, even if the Greek script on the coins did not convey a meaningful text. The same goes for the amulet lamella with its illegible letters: The letters had purely symbolic value and were not intended to communicate textual information. The transmission, adaption and appropriation of earlier traditions in new cultural and religious environments is an important cultural phenomenon. We may locate the silver lamella from Jerash on the cusp of the transition from Greek to Arabic as the dominant language: the very attempt to write a protective amulet in (pseudo-)Arabic, underlines that Arabic was coming to appear as an authoritative religious language worth imitating even if one did not command the language. The lamella was produced in a situation of transformation between Byzantine and Early Islamic culture, and it is thus an impressive witness to the cultural and religious dynamics of this period.

Appendix : Some unintelligible Greek phylacteries on precious metal in the Graeco-Egyptian tradition 37

  • 37 By Richard L. Gordon
  • 38 Kotansky 1991, p. 107.
  • 39 Fernández Nieto 2010. These are written on bronze or stone.
  • 40 I count about 35 ‘activated’ amulets on papyrus in PGrMag and SupplMag, discounting all the prescr (...)
  • 41 On these texts on precious or semi-precious (i.e. copper, bronze) metal, see briefly Bevilacqua et (...)

25In the Graeco-Egyptian tradition phylacteries are texts written out by a practitioner with the aim of protecting a client from specific types of harm. 38 Except for texts against hail and other agrarian ills, which were typically buried on the farmer’s land, 39 the great majority of such phylacteries must always have been written on papyrus or other perishable materials, such as animal-skin/parchment. 40 Those written on precious sheet-metal, like the Jerash lamella, were relatively expensive and thus restricted to the few who could afford them. 41

  • 42 Kotansky 1994. He had apparently intended to collect analogous texts on other Textträger, such as (...)
  • 43 I have made no careful search, but am aware of SEG 30 (1980) 1794 (unknown provenance; gold lamell (...)
  • 44 Amulet against eye-trouble + magical words all in Latin: Kotansky 1994, no. 31; Latin curse with G (...)
  • 45 See the table in Kotansky 1994, xviii. He dates nos. 39, 28, 36 and 51 to between 1st cent. bc and (...)
  • 46 Erotic: AE 2010: 597; to aid conception: Kotansky 1994, no. 61; aggressive: nos. 24, 62; astrologi (...)

26Kotansky’s commented edition of these more expensive texts, published in 1994, contains 68 texts, 36 of which were found in Greek-speaking areas of the Empire, mainly Greater Syria and Greece. 42 22 of these are on silver foil, virtually all the remainder on gold. Since then at least 20 other examples have been published, a few in controlled excavation. 43 Almost all, however, wherever found, are written in Greek. 44 The great majority of them, as one would expect, date from between 2nd and 5th-6th cent. ce45 About one third of Kotansky’s texts are fragmentary, abraded or frankly incompetent. Although one can find a variety of aims, even erotic and aggressive, in one case even an astrological scheme, the most typical purpose is medicinal/curative, including relief from migraine, epilepsy, tertian and quartan fevers, ophthalmic or gynaecological troubles, daemonic/demonic assault, anxieties or nightmares. 46 We can be fairly sure that such an aim inspired the Jerash amulet even if the practitioner’s communicative strategy remained entirely implicit.

  • 47 Charaktêres reject the limitation of script to a fixed set of forms but claim to retain a communic (...)
  • 48 Bevilacqua 2003, p. 118-125 = SEG 53 (2003) 1105.

27This strategy we may term ‘unrecuperability’. As pointed out above, two obvious possibilities are that the writer was illiterate in Arabic, but knew something of the genre, or that he was using a pseudo-script as well as charaktêres47 It seems however worth entertaining a slightly different possibility, namely the deliberate invention of unrecuperable ‘words’ as a means of claiming the authority to intervene in a specific situation. Such texts obey some of the rules in writing natural languages, such as acceptance of paratactic/linear sequential succession of graphic marks, use of signs within a recognised system of writing, an implied (in this case, divine) reader capable of deciphering the signs, but not others, specifically the requirement to employ a shared lexis. A small number of late-Roman phylacteries on precious-metal lamellae have been published that choose precisely this option. The earliest of these (3rd cent. ce), on silver, was once in the Museo Kircheriano in Rome, and its provenance is unknown, but probably Italian. 48 It is evidently an address to a divine power but consists entirely of two columns of letter-sequences, of which I cite those in the right-hand column (B):

Charaktêres
ευστεφσ charaktêres λεγε charaktêres
charaktêres ηβρυωηουακτεινοβραωθθηνωρλαιλαμ
ψσεμεσειλαμοειπιτουκοσμουπυροβουλοσερμαναι,
ερεννουμενενχουχειουθθχανουβηθραναερμανει
αερειναβιωχβιωθχαρνουερινφνιαχνηαχμηφαναχ
[--] σπροκκλα

  • 49 Bevilacqua 2003, p. 121-123 does what she can with these allusions/evocations. I am less convinced (...)

28In this case, the deliberate intention of combining a suggestion of recuperability by including standard or conceivable Greek (λεγε, ακτεινο, του κοσμου, πυροβουλο-, αερει, αρνου) with recognisable voces magicae (which are, for practitioners of Graeco-Egyptian magic ‘standard semi-unrecuperables’) vies with the desire to imitate or evoke an exotic or even occult language full of consonantal collocations impossible in Greek, but nevertheless of high phonetic expressivity. 49 The idea of a communicative system is retained yet simultaneously subverted —the paradox familiar from Dada— in the hope of conveying mastery of the ineffable. To convey this mastery, the writer was prepared to pay the price implicitly required, the virtually total abandonment of natural language.

  • 50 Kotansky 1994, no. 55 (gold, 3rd-4th cent. ce), to which he gives the title ‘Magical Logos’.
  • 51 Kotansky 1994, no. 49, § A 1-12 (4th-5th cent. ce).
  • 52 Kotansky 1994, no. 50 (gold), with a tentative reading beginning: αβη ….ιυροι …βερβερ εχ βε ǀ βε . (...)

29Curiously enough, however, three of the relevant cases were found in the area of Roman Syria and date from the late-Roman period (late 3rd-5th cent. ce), which brings them somewhat closer to our Umayyad text. They may thus suggest the existence of a thin tradition in this area that valued the radical choice of total abandonment of natural language. Two of these, one from near Haifa, Israel and the other from Emesa (Homs), nevertheless employ ordering devices to convey the aim of amuletic protection. In the first, the initial 8 lines of text, beginning .δ.ενια ραβουμα ǀ ραυβαυ ιαθαυ ǀ .ασλα ζανβεαν …, each ‘word’ carefully separated from the previous one, is followed by a roughly-drawn ouroboros enclosing a columnar sequence of voces, four of which clearly begin with Ζ; the lower half of the snake is filled with traces of further letters. 50 The second, on a long thin strip of silver foil, divides its columnar sequences of individual ‘words’ into six clearly delineated ‘boxes’. The penultimate box also contains at least three charaktêres. Yet none of the ‘words’ bears any resemblance to standard divine names, and seem simply to be voces magicae, albeit without much similarity to those in the Graeco-Egyptian tradition; those in the first ‘box’ read: αγεοηλε ǀ κωφαζε ǀ βολκοσε ǀ ναθμουθ ǀ θειγος ευι ǀ ταμηλσεφ ǀ αστραχε ǀ λαμπσαρε ǀ σκιοφωρφ ǀ ηυκταζη ǀ οκταζειν ǀ αμφαυνε. 51 The familiarity shown here with Greek syllabic formation is obvious, and it too implies that a communicative intention lies behind the unrecuperable ‘words’. The final example, from a tomb in Heliopolis (Baʿalbek), of the same period, cannot be properly read because no modern photograph exists, but seems to consist entirely of unfamiliar voces distributed in ten lines, with the notion of communicative intent once again conveyed by the adherence to linear structure and the division of the text into groups of letters recalling words. 52

  • 53 Cf. Mulsow 2012, p. 11-36.

30These four texts may provide a parallel for an amuletic text that deliberately decides to forego recuperability into a natural language in favour of a claim to mastery of ineffability. The communicative act that survives is thus the claim to special or higher knowledge. This leads on to a final point, namely that, although we may term such efforts ‘magic’, they are better understood as claims by ritual specialists, some inside, some outside the ranks of denominated ‘priests’, to be able, through their access to restricted knowledge, to open up the possibility of direct contact with the other world in cases of urgency, desperation or crisis, subjectively defined by clients. Written amulets of this type are thus always to be seen in implicit competition with other types of symbolic aid available in a given social order, just as the knowledge required to create them is always contested and ‘precarious’. 53

The authors would like to thank the following funding bodies for generously supporting our research: the Carlsberg Foundation, The Danish National Research Foundation (DNRF 119), Deutsche Forschungsgemeinschaft (DFG) and H. P. Hjerl Hansen Mindefondet for Dansk Palæstinaforskning. Thanks also go to the Danish-German Jerash Northwest Quarter Project team. We would like to thank the Department of Antiquities (DoA) in Amman and in Jerash without whose continuous support our work could not be carried out. The authors would also like to thank the following people for their comments and input on the inscribed text: Gideon Bohak, John Healey, Ahmad al-Jallad, Christa Müller-Kessler and Shaul Shaked. We also thank Henry Weber and Sandra Engels for dedicating time to training John Møller Larsen in using the computer software as well as Henry Weber for offering input on issues concerning the software and its application. Thanks also go to Gry Barfod (Aarhus University) and Sarah Roeske (UCDavis) for help with electron microprobe analysis and interpretation of the results and to Greg Baxter at UCDavis thin section lab. Further help and advice was given by Philip Ebeling, Helmut Franke, Markus Mayerhöfer, Dirk Neuber, Uwe Peltz and Christof Reinhart.

Haut de page

Bibliographie

Aanavi (D.) 1968 « ‘Pseudo-inscriptions’ in Islamic Art », The Metropolitan Museum of Art Bulletin (New Series) 26/9, p. 353-358.

Album (S.) & Goodwin (T.) 2002 Sylloge of Islamic Coins in the Ashmolean. I. The pre-reform coinage of the early Islamic period, Londres.

Amadasi (M.G.) & Bevilacqua (G.) 2004 « Filatterio greco-aramaico da Roma », MedAnt 7.2, p. 711-725.

Andrade (N. J.) 2013 Syrian Identity in the Greco-Roman World, Cambridge.

Barfod (G. H.), Larsen (J. M.), Lichtenberger (A.) & Raja (R.) 2015 « Revealing text in a complexly rolled silver scroll from Jerash with computed tomography and advanced imaging software », Scientific Reports 5:17765, 2015 (http://www.nature.com/articles/srep17765 [03/05/2016]).

Bauzou (Th.) et al. (éd.) 1998 Fouilles de Khirbet es-Samra en Jordanie, I, La voie romaine. Le cimetière. Les documents épigraphiques, Turnhout.

Bevilacqua (G.) 2003 « Iscrizioni greche magiche di Roma: alcuni espressioni cultuali », Studi epigrafi e linguistici sul Vicino Oriente antico 20, p. 115-130.

Bevilacqua (G.) et al. 2010 Scrittura e magia. Un repertorio di oggetti iscritti della magia greco-romana (Opuscula Epigraphica 12), Rome.

Bohak (G.) 2008 Ancient Jewish Magic: A History, Cambridge.

Canaan (T.) 1937 « The Decipherment of Arabic Talismans [part 1] », Berytus Archaeological Studies 4, p. 69-110.

Clark (V. A.) & Bowsher (J.) 1986 « A Note on Soundings in the Northwestern Quarter of Jerash », Zayadine 1986, p. 343-349.

Coulon (J.-C.) 2010 « Les objets magiques : un indice d’évolution culturelle ? Les documents magiques siciliens entre Byzance et l’Islam », A. Nef & V. Prigent (éd.), La Sicile de Byzance à l’Islam, Paris, p. 95-112.

SupplMag. R. W. Daniel & F. Maltomini (éd.), Supplementum Magicum (Papyrologica Coloniensia 16, 1-2), Opladen, 1990-1992.

Desreumaux (A.) 1998a « Introduction à l’histoire des documents araméens melkites : l’invention du christo-palestinien », Bauzou et al. 1998, p. 3-18.

Desreumaux (A.) 1998b « L’écriture des inscriptions araméennes de Samra », Bauzou et al. 1998, p. 511-522.

Faraone (C. A.) & Kotansky (R.) 1988 « An inscribed gold phylactery in Stamford, Connecticut », ZPE 75, p. 257-266.

Fernández Nieto (F. J.) 2010 « A Visigothic charm from Asturias and the Classical tradition of phylacteries against hail », R. L. Gordon & F. Marco Simón (éd.), Magical Practice in the Latin West: Papers from the International Conference held at the University of Zaragoza, 30 Sept. – 1st Oct. 2005 (RGRW 168), Leyde, p. 551-599.

Foss (C.) 2008 Arab-Byzantine Coins. An introduction, with a catalogue of the Dumbarton Oaks Collection, Washington.

Gawlikowski (M.) 1986 « A residential area by the South Decumanus », Zayadine 1986, p. 107-136.

Gawlikowski (M.) 1995 « Arab Lamp-Makers in Jarash, Christian and Muslim », SHAJ 5, p. 669-672.

Gawlikowski (M.) 2004 « Jerash in Early Islamic Times », Oriente Moderno 23.2, p. 469-476.

Gelzer (T.), Lurje (M.) & Schäublin (C.) 1999 Lamella Bernensis: Ein spätantikes Goldamulett mit christlichem Exorzismus (Beiträge zur Altertumskunde 124), Stuttgart/Leipzig.

Gordon (R. L.) 2011 « Signa nova et inaudita: The Theory and Practice of Invented Signs in Graeco-Egyptian Magical Texts », MHNH: revista internacional de investigación sobre magia y astrología antiguas 11, p. 15-45.

Gzella (H.) 2015 A cultural history of Aramaic: from the beginnings to the advent of Islam, Leyde/Boston.

Healey (J. F.) 2000 « The Early History of the Syriac Script: a Reassessment », JSS 45, p. 55-67.

Kalaitzoglou (G.), Kniess (R.), Lichtenberger (A.) et al. 2012 « Report on a geophysical prospection of the Northwest Quarter of Gerasa/Jarash 2011 », ADAJ 56, p. 79-90.

Kalaitzoglou (G.), Lichtenberger (A.) & Raja (R.) forthcoming « The Danish-German Northwest Quarter Project 2014. Preliminary field report », ADAJ 59.

Kalus (L.) 1981 Catalogue des cachets, bulles et talismans islamiques, Paris.

Kalus (L.) 1986 Catalogue of Islamic seals and talismans, Oxford / New York.

Kennedy (D.) 2007 Gerasa in the Decapolis: A “Virtual Island” in Northwest Jordan, Londres.

Kotansky (R. D.) 1991 « Incantations and prayers for salvation on inscribed Greek amulets », C. A. Faraone & D. Obbink (éd.), Hiera magika: Ancient Greek magic and religion, New York, p. 107-137.

Kotansky (R. D.) 1994 Greek Magical Amulets: The Inscribed Gold, Silver, Copper, and Bronze Lamellae, 1: Published Texts Of Known Provenance. Text and Commentary (Papyrologica Coloniensia 22.1), Opladen.

Kraeling (C. H.) éd. 1938 Gerasa. City of the Decapolis, New Haven.

Lepaon (T.) 2011 « Un nouveau plan pour Jarash/Gerasa (Jordanie) », ADAJ 55, p. 409-420.

Lichtenberger (A.) 2003 Kulte und Kultur der Dekapolis: Untersuchungen zu numismatischen, archäologischen und epigraphischen Zeugnissen, Wiesbaden.

Lichtenberger (A.) & Raja (R.) 2012 « Preliminary report of the first season of the Danish-German Northwest Quarter Project », ADAJ 56, p. 231-240.

Lichtenberger (A.) & Raja (R.) 2015a « New Archaeological Research in the Northwest Quarter of Jerash and its implications for the urban development of Roman Gerasa », AJA 119/4, p. 483-500.

Lichtenberger (A.) & Raja (R.) 2015b « A Hoard of Byzantine and Arab-Byzantine Coins from Jerash », NumChron 175, p. 299-308.

Lichtenberger (A.) & Raja (R.) 2016 « Jerash in the Middle Islamic period. Connecting texts and archaeology through new evidence from the Northwest Quarter », ZDPV 132, p. 63-81.

Lichtenberger (A.) & Raja (R.) et al. 2016 « A newly excavated private house in Jerash. Reconsidering aspects of continuity and change in material culture from Late Antiquity to the early Islamic period », J. C. Magalhães de Oliveira, Potestas populi (AntTard 24), Paris.

Lichtenberger (A.), Raja (R.) & Sørensen (A. H.) forthcoming « Preliminary registration report of the fourth season of the Danish-German Jerash Northwest Quarter Project 2014 », ADAJ 59.

Maddison (F.) & Savage-Smith (E.) éd. 1997 Science, Tools & Magic. I. Body and Spirit, Mapping the Universe (The Nasser D. Khalili Collection of Islamic Art XII), Londres.

Mershen (B.) 1983 Untersuchungen zum Schriftamulett und seinem Gebrauch in Jordanien. Dargestellt am Beispiel der Stadt Jarash und ihres ländlichen Umfeldes. Dissertation Mainz University.

Mocella (V.), Brun (E.), Ferrero (C.) & Delattre (D.) 2015 « Revealing letters in rolled Herculaneum papyri by X-ray phase-contrast imaging », Nature Communications 6:5895. doi: 10.1038/ncomms6895 [04/05/2016].

Müller-Kessler (C.) 2005 Die Zauberschalentexte in der Hilprecht-Sammlung, Jena, und weitere Nippur-Texte anderer Sammlungen, Wiesbaden.

Mulsow (M.) 2012 Prekäres Wissen: Eine andere Ideengeschichte der frühen Neuzeit, Francfort.

Naveh (J.) & Shaked (S.) 1987 Amulets and magic bowls: Aramaic incantations of Late Antiquity, Jérusalem, 2nde éd.

Naveh (J.) & Shaked (S.) 1993 Magic Spells and Formulae: Aramaic Incantations of Late Antiquity, Jérusalem.

Pierobon (R.) 1983 « Gugliemo di Tiro e il castrum di Gerasa », Prospettiva Settanta 1, p. 8-13.

Pierobon (R.) 1983-1984 « Gerasa in Archaeological Historiography », Mesopotamia 18-19, p. 13-35.

Porter (V.) 2011 Arabic and Persian Seals and Amulets in the British Museum, Londres.

PGrMag K. Preisendanz (éd.), Papyri Graecae Magicae. Die griechischen Zauberpapyri 2, Stuttgart, 1973-1974, éd. rév. par A. Henrichs.

Raja (R.) 2012 Urban Development and Regional Identity in the Eastern Roman Provinces, 50 bc – ad 250: Aphrodisias, Ephesos, Athens, Gerasa, Copenhague.

Savage-Smith (E.) 1997a « Amulets and related talismanic objects », Maddison & Savage-Smith 1997, p. 132-145.

Savage-Smith (E.) 1997b « Magic and Islam », Maddison & Savage-Smith 1997, p. 59-71.

Schaefer (K. R.) 2006 Enigmatic charms. Medieval Arabic Block Printed Amulets in American and European Libraries and Museums (HdO Section 1, The Near and Middle East 82), Leyde/Boston.

Schienerl (P. W.) 1984 « Der Ursprung und die Entwicklung von Amulettbehältnissen in der antiken Welt », AntWelt 15, p. 45-54.

Seales (W. B.) & Delattre (D.) 2013 « Virtual unrolling of carbonized Herculaneum lamellas: Research Status (2007–2012) », Cronache Ercolanesi 43, p. 191-208.

Segal (J. B.) 2000 Catalogue of the Aramaic and Mandaic Incantation Bowls in the British Museum, Londres.

Seigne (J.) 1989 « History of Exploration at Jerash, The Sanctuary of Zeus », D. Homes-Fredericq & J. B. Hennessy (éd.), Archaeology of Jordan II/1, Field Reports, Surveys and Sites (A-K), Louvain, p. 319-323.

Tholbecq (L.) 1997-1998 « Une installation d’époque islamique dans le sanctuaire de Zeus de Jérash (Jordanie) : La céramique », ARAM 9-10, p. 153-179.

Tomlin (R. S. O.) 2014 « “Drive away the cloud of plague”. A Greek amulet from the City of London », R. Collins & F. McIntosh (éd.), Life in the Limes: Studies for L. Allason-Jones, Oxford, p. 197-205.

Tsafrir (Y.) & Foerster (G.) 1992 « The dating of the ‘Earthquake of the Sabbatical Year’ of 749 ce in Palestine », BSOAS 55(2), p. 231-235.

Walmsley (A. G.) 2008 « The Middle Islamic and Crusader Periods », R. B. Adams (éd.), Jordan. An Archaeological Reader, Londres, p. 495-537.

Walmsley (A. G.) 2011 « Trends in the urban history of Eastern Palaestina Secunda during the Late Antique – Early Islamic Transition », A. Borrut et al. (éd.), Le Proche-Orient de Justinien aux Abbassides : peuplement et dynamiques spatiales (Actes du colloque « Continuités de l’occupation entre les periods byzantine et abbaside au Proche-Orient, viie-ixe siècles », Paris, 18-20 octobre 2007, Turnhout, p.  271-284.

Walmsley (A. G.), Blanke (L.), Damgaard (K.) et al. 2008 « A mosque, shops and bath in central Jarash: the 2007 season of the Islamic Jarash Project », ADAJ 52, p. 109-138.

Winkler (H. A.) 1930 Siegel und Charaktere in der muhammedanischen Zauberei, Berlin.

Woltering (R.) 2015 « Pseudo-Arabic », Encyclopedia of Arabic Language and Linguistics, http://referenceworks.brillonline.com/entries/encyclopedia-of-arabic-language-and-linguistics/pseudo-arabic-EALL_SIM_000025

Zayadine (F.) éd. 1986 Jerash Archaeological Project, 1, 1981-1983, Amman.

Haut de page

Notes

1 For introductory literature on Gerasa and further references see: Kraeling 1938; Zayadine 1986; Lichtenberger 2003, p. 191-243; Kennedy 2007; Raja 2012, p. 137-189. Andrade 2013, p. 160-169 presents the most recent summary of the history of the city. Also see the collection of articles in Syria 66, 1989, p. 1-261, which relate to the history and archaeology of Gerasa. See the recent article by Lichtenberger & Raja 2015a for the first three campaigns in Jerash by the Danish-German Northwest Quarter Project and further references as well as new considerations on the implications of the results for the overall development of Gerasa in the Roman period.

2 On this earthquake see Tsafrir & Foerster 1992.

3 Walmsley 2008 and Walmsley 2011 as well as Walmsley et al. 2008 for considerations on the Abbasid material.

4 Lichtenberger & Raja 2016 for a presentation of the Mamluk evidence from the Northwest Quarter in Jerash as well as a summary of the evidence published by other missions working in Jerash. Furthermore Gawlikowski 1986, Pierobon 1983 and Pierobon 1983-1984, Seigne 1989 as well as Tholbecq 1997-1998 remain essential on the Mamluk period material.

5 Clarke & Bowsher 1986 for a short contribution on soundings in the Northwest Quarter, which yielded evidence from the 3rd cent. ce.

6 Lichtenberger & Raja 2012 and Kalaitzoglou, Kniess, Lichtenberger et al. 2012 for publications on the first season of examinations undertaken by the Danish-German Northwest Quarter Project.

7 Lichtenberger & Raja 2015a.

8 Lichtenberger & Raja 2016.

9 For a more detailed description of the building see Lichtenberger & Raja et al. 2016 as well as Kalaitzoglou, Lichtenberger & Raja forthcoming and Lichtenberger, Raja & Sørensen forthcoming.

10 Lichtenberger & Raja 2015b.

11 The lamella and its case have the find number: J14_Kh_3_498a_M113.

12 The technical aspects of this process and the analysis of the metals are described in detail by Barfod, Larsen, Lichtenberger & Raja 2015.

13 Cf. for such containers Schienerl 1984.

14 See Seales & Delattre 2013 as well as Mocella, Brun, Ferrero & Delattre 2015.

15 Anyhow, it is not possible to identify the lamella script with Mandaic, personal communication from Christa Müller-Kessler.

16 Gzella 2015, p. 281-329 provides a useful overview of Aramaic in this region in Late Antiquity.

17 An overview of the locations of the epigraphic material is found in Desreumaux 1998a, p. 10-11. Discussion and analysis of the script are found in Desreumaux 1998a, p. 8-9 and 1998b.

18 Contrary to a widespread earlier view in research, the more cursive serto script was already present alongside Estrangela in the first centuries ce and not a late derivation of Estrangela, see Healey 2000.

19 A chart of the Christian Palestinian script is found in Desreumaux 1998b, p. 513.

20 For examples of amulets without diacritics, see for instance Savage-Smith 1997a, no. 58 (p. 132), 70-72 (p. 136-137). There are also several examples of printed amulets without diacritics in Schaefer 2006.

21 Canaan 1937, p. 81-83.

22 Kalus 1981, no. 2.10 (p. 98 and pl. 16).

23 Woltering 2015.

24 Aanavi 1968, p. 353-358.

25 Savage-Smith 1997a notes that “Pseudo-Kufic writing was (…) frequently employed for talismanic decoration” (p. 134) and reproduces a specimen: no. 60 (p. 132); several specimens of such undeciphered Kufic can be seen in Porter 2011, p. 177-180. Pseudo-script is also found in Kalus 1981, no. 1.36 (p. 89-90 and pl. 14 —however, the author notes that undoubtedly it is of recent, and perhaps Western, production); Kalus 1986, no. 1.88 (p. 91).

26 Segal 2000, no. 121Ps-142Ps (p. 150 and Pl. 138-159); Müller-Kessler 2005, p. 5-6, 244-257.

27 Kotansky 1994, no. 1 (p. 1-2), 3 (Latin, p. 13-15), 12 (p. 54-57), 14 (p. 72), 18 (p. 81-88), 41 (p. 220-231), 43 (p. 234), 45 (p. 236-238), 46 (p. 239-244), 57 (p. 326-330), 62 (p. 369-373), 66 (p. 379-382).

28 Naveh & Shaked 1987, no. 2 (p. 44-49 and pl. 2), 4 (p. 54-60 and pl. 4), 5 (p. 60-63), 8 (p. 78-81 and pl. 7), 14 (p. 101-105 and pl. 12); Naveh & Shaked 1993, no. 18 (p. 57-60 and pl. 4), 19 (p. 60-66 and pl. 6-7), 20 (p. 67-68 and pl. 5a), 21 (p. 68-72 and pl. 8), 25 (p. 85-87 and pl. 11), 27 (p. 91-95 and pl. 13), 30 (p. 101-105 and pl. 16), 31 (p. 105-107 and pl. 17), 32 (p. 107-109 and pl. 18).

29 Naveh & Shaked 1987, bowl 1 (p. 124-132 and pl. 14-15); Naveh & Shaked 1993, bowl 26 (p. 139-142 and pl. 31).

30 Gordon 2011. A recent study (Coulon 2010) emphasizes the importance of the Greek-Byzantine tradition for the introduction of ‘ring-letters’ in early Islamic magic.

31 Bohak 2008, p. 270-274; Winkler 1930, p. 150-167.

32 Walmsley et al. 2008, p. 109; see also p. 119-122, p. 124-126.

33 Gawlikowski 1995 et 2004, p. 470.

34 Savage-Smith 1997b, p. 60.

35 See the study on such modern amulets from Jerash: Mershen 1983.

36 See the introductions by Foss 2008 and Album & Goodwin 2002.

37 By Richard L. Gordon

38 Kotansky 1991, p. 107.

39 Fernández Nieto 2010. These are written on bronze or stone.

40 I count about 35 ‘activated’ amulets on papyrus in PGrMag and SupplMag, discounting all the prescriptions for phylacteries in the receptaries and the Christian examples. A Jewish amulet in the Cairo Geniza to be written on deer-skin: Bohak 2008, p. 236.

41 On these texts on precious or semi-precious (i.e. copper, bronze) metal, see briefly Bevilacqua et al. 2010, p. 23-32. The Graeco-Egyptian magical papyri rarely prescribe the use of lamellae of gold or silver, but see e.g. PGrMag III 297 (divination); X 26 (to allay anger).

42 Kotansky 1994. He had apparently intended to collect analogous texts on other Textträger, such as tin/pewter, lead, bone, ostraka or wood, but this volume has never appeared. In the relatively few cases where the provenience is known, they were found in tombs, i.e. had been buried with the deceased: Gelzer, Lurje & Schäublin 1999, p. 5.

43 I have made no careful search, but am aware of SEG 30 (1980) 1794 (unknown provenance; gold lamella; epilepsy, 3rd cent. ce); ZPE 75 (1988) p. 257-266 (unknown provenance, gold lamella, general protection, mid-4th cent. ce); SEG 41 (1991) 1494 (Cyprus, lead, invoking Gabriel, Rephael); 42 (1992) 1582 (Tell el-Amarna, silver, 4th cent. ce); 45 (1995) 272 = 53 (2003) 1110 (unknown, gold); AE 1999: 313 (Rome, silver, epilepsy [ ?], 4th-5th cent. ce); SEG 49 (1999) 2383 (‘lamella Bernensis’, silver, Christian: highly accomplished); ZPE 156 (2000) p. 103-108 (Sammardenchia, Udine, gold, σῷζε, σῷζε); AE 2002: 484 = 2006: 446 (Forum Fulvii, Reg. IX, gold, woman’s tomb, migraine); 2002: 565 (Sarcedo, Reg. X., Latin, 4th cent. ce); 2002: 577 (Arco, Lake Garda, silver, late 2nd/early 3rd cent. ce); 53 (2003) 1105 (unknown, silver); 2004: 661 (Comiso, Sicily, gold, Jewish influence, 5th cent. ce); 2005: 231 (Rome, man’s tomb, gold); SEG 57 (2007) nos. 923/924 (Gnathia, Campania, 2 gold lamellae, 4th cent. ce); ibid. 2065 (‘Christian’, tin (?), with ouroboros, late Roman); AE 2008: 778 (Oxford, gold lamella rolled, unusual amulet for a safe birth, 250-350 ce); AE 2010: 597 (Riva del Garda, necropolis, gold lamella, rolled up; amatory, 4th cent. ce); 2011: 265-266 (S. Maria Capua Vetere, necropolis, 2 silver amulets against afflictions, late 4th-5th cent. ce); Britannia 24 (2013) p. 390-391 no. 21 = Tomlin 2014 (from the Thames in the City of London, pewter, partly metrical, against plague [λυμοῦ, l. 5], probably Antonine); ibid. 46 (2015) p. 406 no. 42 (City of London, pewter, charaktêres only). Faraone & Kotansky 1988, p. 257 n. 2 mention a cache of 19 unpublished silver lamellae (once) owned by the J. P. Getty Museum in Pacific Palisades, possibly from Turkey.

44 Amulet against eye-trouble + magical words all in Latin: Kotansky 1994, no. 31; Latin curse with Greek vowel-sequence: no. 24; one or two Latin words in a Greek matrix: no. 18 (new reading in SEG 57 [2007] 999); Latin in Greek letters + Latin: no. 7 = AE 2002: 1063; voces magicae (?) in Greek, the word Veneris in Latin: AE 2006: 868 (Zülpich, Germ. Inf.; gold lamella wrapped around a gold ring); voces magicae in Latin script: nos. 17, 63; 3 angel names in Latin script: no. 4; Gallo-Latin: no. 8; Gaulish: AE 1993: 1203 = 2007: 988 (v. doubtful); Latin/Greek/Latin in Greek characters: AE 2004: 853 = 2006: 749 (nr. Dereham, Norfolk, gold lamella, niketikon carried in a capsule. Undated but tria nomina). Hebrew and Greek: Kotansky 1994, no. 56 (modern Israel, 3rd-4th cent. ce); Aramaic/Greek: Amadasi & Bevilacqua 2004 (now in Rome, 5th cent. ce); Old Coptic: AE 2003: 1240 (Vindonissa, two lamellae in a capsule, one gold, one silver; found in a woman’s grave of the 5th cent. ce, and evidently in use for 200 years).

45 See the table in Kotansky 1994, xviii. He dates nos. 39, 28, 36 and 51 to between 1st cent. bc and 1st cent. ce. To my mind the strongest case for an early date is no. 36 (Amisus, Pontus, silver), intended to avert a lawsuit, which combines Egyptian and Jewish names of power.

46 Erotic: AE 2010: 597; to aid conception: Kotansky 1994, no. 61; aggressive: nos. 24, 62; astrological: Kotansky 1994, no. 54. Nos. 40, 58 and 60 are for favour and victory (no. 58 in an important hearing before the governor of Arabia). PGrMag IV 2154-55 enjoins hanging an iron lamella round a dying man as an amulet to force him to speak the truth. Kotansky 1994, no. 6 = AE 2002: 1063 (Badenweiler) seems to have been intended to protect three slaves threatened with punishment, e.g. interrogation by the family court, whipping or even torture (Kotansky, however, suggests a trial-context).

47 Charaktêres reject the limitation of script to a fixed set of forms but claim to retain a communicative intention, in that they forge a direct linguistic link with the Other World.

48 Bevilacqua 2003, p. 118-125 = SEG 53 (2003) 1105.

49 Bevilacqua 2003, p. 121-123 does what she can with these allusions/evocations. I am less convinced by her wish to see a divinatory purpose here, connected to Hermes (p. 123-125).

50 Kotansky 1994, no. 55 (gold, 3rd-4th cent. ce), to which he gives the title ‘Magical Logos’.

51 Kotansky 1994, no. 49, § A 1-12 (4th-5th cent. ce).

52 Kotansky 1994, no. 50 (gold), with a tentative reading beginning: αβη ….ιυροι …βερβερ εχ βε ǀ βε .βιορβ.θε μ θυλλε … ..η …..

53 Cf. Mulsow 2012, p. 11-36.

Haut de page

Table des illustrations

Titre Figure 1.
Légende Map of Jerash with Northwest Quarter marked (after Lepaon 2011)
URL http://journals.openedition.org/syria/docannexe/image/4656/img-1.jpg
Fichier image/jpeg, 1,2M
Titre Figure 2.
Légende Map of the Northwest Quarter in Jerash with the location of all trenches excavated by the Danish-German Northwest Quarter Project in the period 2011-2015. Trench K is the trench in which the lamella was found.
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-2.jpg
Fichier image/jpeg, 771k
Titre Figure 3.
Légende Orthophoto of trench K
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-3.jpg
Fichier image/jpeg, 779k
Titre Figure 4.
Légende The container while the lamella was still inside
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-4.jpg
Fichier image/jpeg, 828k
Titre Figure 5.
Légende The container and the lamella separated after conservation
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-5.jpg
Fichier image/jpeg, 575k
Titre Figure 6.
Légende Close-up of the container showing the traces of textile fibres
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-6.jpg
Fichier image/jpeg, 932k
Titre Figure 7.
Légende The silver lamella
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-7.jpg
Fichier image/jpeg, 293k
Titre Figure 8.
Légende A slice of the lamella showing how complexly it was rolled before being placed in the container
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-8.jpg
Fichier image/jpeg, 129k
Titre Figure 9.
Légende Illustration showing the location of the slice depicted in fig. 8
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-9.jpg
Fichier image/jpeg, 131k
Titre Figure 10.
Légende Lines 1-8 from the silver lamella. Shown are the back sides, mirrored. “a” and “b” represent upper and lower parts of a line.
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-10.jpg
Fichier image/jpeg, 800k
Titre Figure 11.
Légende Lines 9-17 from the silver lamella. Shown are the back sides, mirrored
Crédits © The Danish-German Northwest Quarter Project
URL http://journals.openedition.org/syria/docannexe/image/4656/img-11.jpg
Fichier image/jpeg, 822k
Haut de page

Pour citer cet article

Référence papier

John Møller Larsen, Achim Lichtenberger, Rubina Raja et Richard L. Gordon, « An Umayyad period magical amulet from a domestic context in Jerash, Jordan », Syria, 93 | 2016, 369-386.

Référence électronique

John Møller Larsen, Achim Lichtenberger, Rubina Raja et Richard L. Gordon, « An Umayyad period magical amulet from a domestic context in Jerash, Jordan », Syria [En ligne], 93 | 2016, mis en ligne le 01 novembre 2018, consulté le 28 mai 2020. URL : http://journals.openedition.org/syria/4656 ; DOI : https://doi.org/10.4000/syria.4656

Haut de page

Auteurs

John Møller Larsen

Institute for Culture and Society, Aarhus University, Denmark

Achim Lichtenberger

Institut für Archäologische WissenschaftenRuhr-Universität Bochum, Germany

Articles du même auteur

Rubina Raja

Classical Art and Archaeology

Articles du même auteur

Richard L. Gordon

Erfurt University, Max Weber Center for Advanced Cultural and Social Studies, Germany

Articles du même auteur

Haut de page

Droits d’auteur

© Presses IFPO

Haut de page
  • Logo IFPO
  • Logo CNRS - Institut des sciences humaines et sociales
  • Logo Maison Archéologie & Ethnologie
  • Logo ERIH PLUS | NSD
  • OpenEdition Journals