34

Introduction

Although Petra, the ancient capital of theNabataean Kingdom, located in southernJordan, has always attracted the attention ofhistorians and archaeologists, the Byzantineperiod (4th-early 7th century AD) there waspoorly known and understood. However,the 1990s have witnessed a considerableexpansion of archaeological activities inPetra and its vicinity. This recent interest inthe Byzantine period also includes theinvestigations of Christian monuments inthe Petra area.2 In this context, of particularimportance is Jabal an-Nabi Harûn (themountain of the Prophet Aaron) located ca 5km SW of Petra, and often referred to as theBiblical Mount Hôr by the early explorersof the area. According to the Jewish,Christian and Muslim traditions, themountain is considered to be the place ofburial of Moses’ brother Aaron. Currently,the peak of the mountain is occupied by the14th century Muslim shrine (weli)3, with asarcophagus believed to contain Aaron’sremains. Furthermore, there is an extensive,ruined architectural complex located ca 70m below and ca 150 m to the west of thepeak with the weli, on a wide plateau of themountain, and at ca 1270 m above sea level.The extant historical information, albeitlimited, and the results of the earlyexplorations at the mountain indicated that

a Byzantine monastery should be locatedthere.

Historical Information and PreviousExploration

Eusebius, the 4th century Church historian,mentioned Mount Hôr near Petra.4 SomeByzantine records refer to Mâr Harûn as aplace frequented by monks during their walksaround the Dead Sea during Lent,5 but thelocation of this toponym is uncertain. Animportant information was provided by Al-Mas’ûdî (mid-10th century) who specifiedJabal Hârûn as a holy mountain of theChristians in the possession of the Melkites.6

The place was visited by the Crusaders duringthe expedition of Baldwin I to Transjordan in1100: “...Furthermore we found at the top ofthe mountain the Monastery of St Aaron whereMoses and Aaron were wont to speak withGod. We rejoiced very much to behold a placeso holy and to us unknown...”7 Other Crusaderrecords mentioned the presence of a churchthere.8 The last reference is that by MagisterThetmarus who noted a church and two Greekmonks living there in 1217.9

The mountain was briefly mentioned bythe 19th century explorers, among themBurckhardt, Irby and Mangles, and Palmer.10

Only in the early 20th century, importantexplorations were conducted there by Musil

Arkeologipäivät 2002

The Byzantine monastic / pilgrimage center of St.Aaron near Petra, Jordan1

Zbigniew T. Fiema

35

(1900) and Wiegand (1916). The latterconcentrated on the structure of the weli, datedby the dedicatory inscription to 1363.Wiegand provided a plan and the descriptionof the weli, noting pieces of church-related,marble furnishing and opus sectile pavement,either reused in the weli construction, ordiscarded nearby. He proposed that theextensive masonry upon which the weli wasconstructed represented the remains of a largememorial church built on the central plan andassociated with side rooms. As for the ruinedcomplex on the plateau below the mountain’ssummit, some explorers admitted a possibilitythat the ruins related to a Byzantinemonastery. However, it was the 1991exploration by Russell, Peterman and Schick,which resulted in a first sketch-plan of thatsite, its description, and a proposition that theruins should probably be identified with themonastery of Saint Aaron mentioned inhistorical sources.12

This proposition was further reinforcedby the information provided by the PetraPapyri. The 1993 discovery of thecarbonized archive of the 6th century GreekByzantine papyri in a room adjacent to thePetra church has opened an entirely newavenue of research on Byzantine Petra. Thetexts are mainly legal documents concerningtransactions and registrations of property,and they also mention local towns,churches, and dwellings, as well as theagricultural hinterland of Petra.13 PapyrusPetra inv. 6 (Papyrus Petra Daniel C. andNancy E. Gamber) dated to 15 June, A.D.573, also mentions “the House of our Lordthe Saint High-Priest Aaron” outside of thecity of Petra.14 This institution is mentionedin the will of a person, as one of the twobeneficiaries in the event of his death(donatio propter mortem). Because of theoccurrence of the terms Hagios Oikos, inGreek, and Domus, in Latin, and of the titleof Hegoumenos, the papyrus almostcertainly refers to a monastery of SaintAaron near Petra.

The Finnish Jabal Harûn Project -Objectives and Results

The combination of this information withthe aforementioned religious traditionassociated with Jabal Harûn, and the results ofthe early exploration in the area, wouldstrongly suggest that the architectural remainson the high plateau, which were otherwiserecognized as remains of a monastic complex,can indeed be identified as the Monastery ofSaint Aaron. However, the ultimateconfirmation of this hypothesis could comeonly through the archaeological excavationsof the ruined complex. To this effect, theFinnish Jabal Harûn Project (FJHP) began thecomprehensive investigation of the site and itsenvirons in 1997. The project is directed byProf. Jaakko Frösén, University of Helsinki,and sponsored by the Academy of Finland.The participants include archaeologists andcartographers from the University of Helsinkiand the Helsinki University of Technology,respectively, as well as archaeologists andconservators from the USA, Sweden, andItaly.

The FJHP is designed as a multi-seasonand interdisciplinary investigation. The mostimportant goal of the Project is the study of thespatial and temporal variations in humanoccupation in the area of Jabal Harûn, with thespecial emphasis on the extent and nature ofoccupation at the site recognized as aByzantine monastery. Furthermore, theProject investigates patterns of humanadaptation in the area, the palaeo-environmental variations, aspects of land-use,ancient agriculture and resource exploitation.Ultimately, the Project will address the issueof the relationship between the Jabal Harûnarea and the city of Petra, from the Nabataeanthrough the Early Islamic periods. To meetthese objectives the project utilizesarchaeological excavations and survey,cartographical fieldwork and research,architectural studies, geoenvironmental

36

exploration, paleoethnobotanical andpaleozoological research, and ethno-archaeological survey (Fig. 1).

So far the FJHP conducted anarchaeological reconnaissance (1997) and fivefull excavation seasons (1998-2002) at the siteof the ruined complex at the high plateau ofthe mountain (Figs 2 and 3).15 The excavationspartially exposed a large basilican church anda chapel, and some auxiliary structures androoms. The research on the data and findsprovided by the fieldwork indicates that thecomplex, in addition to its monastic function,

had most probably also served as a pilgrimagecenter dedicated to the veneration of St.Aaron. This monastic-pilgrimage centerappears to have existed between the later 5thand the early 8th century AD., if not later.

The Site Description

The site of the complex measures ca 62 mN-S x 48 m E-W. Generally, the complex canbe divided into four main components orwings situated around three courts. The

Fig. 1. Visualization of the 3D model of the Jabal Harûn mountain. The excavation site is visibleas a rectangular, slightly elevated countour in the central-left (close to the edge of the plateau).Drawing: R. Karjalainen.

37

Fig. 2. The plan of the Byzantine monastic/pilgrimage center at Jabal Harûn - the results of the2002 campaign are not included. Drawing: K. Koistinen and V. Putkonen.

38

asymmetrical location of the western wing vs.the eastern, southern and northern parts of thecomplex may indicate that the former is notcontemporary with the other threecomponents. The central location is occupied

by the church and a chapel, the formerpreceded by a narthex and then by a court witha cistern cut in the bedrock. Farther west is thewestern wing - a long, N-S oriented compositestructure which consists of separate rooms

(Fig. 4). One of theserooms, recentlyexcavated, is a solidstone-filled compart-ment, ca 6 x 5 m,which appears as apodium or platform.The structure isassociated with two

Fig. 3. The apprearance of the excavation site after the 2002 campaign. View from the weli.Photo: Z.T. Fiema.

Fig. 4. The mon-umental compositebuilding located in thewestern side of themonastic site. Photo:Z.T. Fiema.

39

staircases. Directlynorth is a room withthree arches, while asimilar room islocated also to thesouth of the stone-filled room. A large,well-built stone scarp(glacis) was uncover-ed on the slope west ofthe structure. North ofthe chapel is a large court surrounded on threesides by 14 rooms of a substantial size. Bothconstruction techniques and the material usedthere appear somewhat inferior to these usedin the construction of the church. The plan andappearance of the northern part of the complexis not unlike that of a caravanserai. Probablythe court and the rooms functioned as a hostelfor pilgrims. Two rooms were excavated thereso far. Both had the multi-phased occupationand they provided an excellent stratigraphicsequence of ceramic and glass material. Southof the church is another court also limited byseries of rooms on its southern side.

In the SW corner of the site a large, two-arched room was excavated. It’s interiorprovided evidence for multi-phasedoccupation which probably had someeconomic significance. A large fragment of arotating grain mill was found there. Directlysouth of this room a water channel and amasonry-built basin were uncovered, the latterprobably a threading floor for the productionof oil or wine (Fig. 5). Both installations werefound covered by strata of ancient trash anddebris. Apparently, this area went out of use ata certain point of time, and it became a generaldisposal place. The latest there was a middenwhich contained large quantities of fish scalesand bones, primarily of Scaridae (parrotfish)and Serranidae (groupers).

Although the main entrances were notexcavated yet it is strongly suspected thatthere were two: one on the western side,north of the room with three arches, andanother one near the SW corner of thecomplex. The court with the cistern appearsto have been the main communication hubof the entire complex, through which onecould proceed from the South to the NorthCourt. The area around the cistern waspaved with large flagstones. Two phases ofpavement were discerned there. Threechannels carried out rainwater from the areaof the church (and under the narthex) intothe cistern. One of these was an elaborateconstruction with capstones and a settlingtank lined up with hydraulic mortar. A fewmeters NE of the cistern, there is amultiroomed structure which featuredseveral phases of occupation and anextensive evidence of remodelling andchanges in function. During the later phasesof existence, the central part of the structurereceived waterproof plaster and thus wasprobably related to some sort of activitiesincluding water storage. Still later, this areabecame a dump of lime slags and acollection point of stone tesserae and glassshards. As in other parts of the complex, theevidence of very late temporary or casualoccupation is abundant here.

Fig. 5. The masonry-built basin in the SWarea of the site. Photo:Z.T. Fiema.

40

Phasing of the Church and the Chapel

Although the relative chronology ofparticular structures at the site is wellestablished, and significant chronologicalindicators (ceramics, lamps, and glass) werefound in well-stratified deposits, it is notpossible at this time to offer an overallchronological sequence for the entire site. Theexcavated structures are often distanced fromeach other. Discrete strata or specificmodifications can be dated there but thegeneral phasing can be proposed only whenthe physical connection between thesestructures is achieved. The exceptions are thechurch and the chapel which form one unit andthe excavations of which provided most of theinformation available so far. Following is apreliminary phasing of the occupation andmodifications of these ecclesiasticalstructures, preceded by short comments on thepre-Byzantine phases at the site.

Nabataean-Roman Phases

Initially, the pre-Byzantine period at thesite was known only from the Nabataeanceramics and few architectural stones -limestone lintels with elaborate mouldings,which were re-used in the church construction.Very similar mouldings were found in somebuildings in Petra, dated to the first centuryAD. The composite structure in the westernwing of the complex appears to be the onlybuilding known so far at the site, whichapparently originated in the pre-Byzantineperiod. Its masonry, consisting of largeashlars, is unlike any other at the site. Also,the evidence of plaster on the exterior,probably an inner layer for stucco decoration,and the overall appearance would all stronglyindicate a monumental Nabataean design. Thefunction of the structure is uncertain, bothduring the Nabataean-Roman period and afterit was incorporated into the Byzantinemonastic complex. But the elaborate character

of the structure suggests something moresubstantial than a simple observation ordefensive tower. The structure is located in avery conspicuous place - on the plateau of themost prominent mountain in the Petra area -and overlooking the area of the Wadi ‘Araba,which may support a sacral character of thestructure. Evidently, such function of thestructure ceased in the Byzantine period.

Phase I: Early Church and the Chapel

Although the church experienced severalphases of remodelling, it is most probable thatits structure belongs to the earliest phase of theByzantine occupation at the site. In this phase,the church was a tripartite, monoapsidalbasilica, internally measuring ca. 22.6 m(max) x 13.6 m, with seven columns in each ofthe two rows. Comparing with the size of thePetra church, dated to the later fifth centuryAD., which is internally ca 23.21 m. long andca 15.35 m. wide,16 the Jabal Harûn church isof close dimensions. Therefore, the ratio of theinner length to inner width, being 3 : 2 for thePetra church, is comparable for the the JabalHarûn church. This length to width ratio isrelatively typical for earlier churches inPalestine (4th-5th century), characterized bylong and narrow aisles.17 The later 5th centurydate for the Jabal Harûn church is alsosupported by the ceramic material recoveredfrom the inner fill of the main walls of thechurch, which was not later than the mid-5thcentury.

The apse, ca 5.2 m long at the chord, wasflanked by two pastophoria, similarly to thePetra church in its early phase. The southpastophorion contained a cupboardinstallation in its northern wall. Marble floorwas laid out throughout the church. Marbleslabs, some recognized as Proconnesianmarble, were apparently taken from thedisused monumental structures in Petra. It isnot certain whether or not the church waspreceded by a simple narthex in this phase.

41

The early, marble-clad, rectangular bema wasunusually narrow but fully contained withinthe nave, as in the Petra church.18 The apsehad a two-tiered synthronon installation,which shows affinities with the five-tieredsynthronon of the Petra church.The clearlypreserved remains of the bishop’s throne inthe Jabal Harûn church were accessed by theset of steps centrally superimposing thesynthronon tiers. However, while the JabalHarûn synthronon is clearly an originalinstallation, and not added later, as in the Petrachurch, the throne appears to have been addedlater (infra). The one-to-three tier synthronontypes are generally better attested before the5th century - an observation which alsosupports the 5th century date for the JabalHarûn church.

So far only the southern half of thechapel’s apse and the western end of thestructure have been excavated. Apparently,the chapel was built in the same time as theearly church, and sharing the wall between

them. The eastern end featured an apseflanked on the south by a high cupboard orcabinet with three shelves. Remains of a floormade of marble slabs have been found in thesounding in the apse. In the proximity of theextant western wall of the chapel, a roughlyoctagonal pit was chiselled out of the bedrockand the cruciform baptismal font with themasonry-built upper part was installed andfurther integrated with the bedrock usingmortar (Fig. 6). The font is small in size: ca0.92 m (N-S) x 0.89 m (E-W) at the opening,and no more than 0.6 m deep. No floorremains associated with this phase were foundin this area of the chapel. Possibly, the fontwas almost entirely sunk below the floor level,but perhaps with an elevated rim.

The font belongs to the cruciform typewhich is usually masonry-built and generallyearlier in date than the monolithic fonts.19 Aclose parallel is the large, canopied cruciformfont found in the baptistery of the Petrachurch. Cruciform fonts were popular in

Fig. 6. The cruciform baptismal font in the chapel. Photo: Z.T. Fiema.

42

southern Palestine and especially in theNegev, e.g., in the East Church at Mampsis,the North Church at Oboda/’Avdat, and theNorth and South Churches at Sobata/Shivta;the first two masonry-built.20 With proposedlater 5th century dating for the beginning ofPhase I of the church and the chapel at JabalHarûn, the baptismal font there should beconsidered, together with that at the PetraChurch, as one of the earliest known structuresof this kind in Jordan.

Baptismal fonts are not uncommon in themonastic context and, in fact, they tend tooccur in the monasteries associated with aholy place or a pilgrimage center, and often ina non-urban location.21 But the location of thefont in the western part of the chapel, and theoverall function of the chapel during Phase Iremain puzzling, even if the western wall ofthe chapel was in that phase located fartherwest. In baptisteries with no apse, font wasusually located toward the eastern end of theroom, as to emulate the relationship betweenthe community and the altar, symbolized bythe font.22 A good example is provided by thecruciform font located near the eastern end ofthe apse-less Old Diaconikon at the Memorialof Moses on Mt. Nebo.23 However, the JabalHarûn chapel in this phase had an apse (and analtar?), yet the font was located far away fromit. Presumably, some, still not exposedinstallations or partitions existed between thefont and the apse of the chapel during Phase I.In fact, the location of the Jabal Harûn fontmay resemble that at the church in HorvatKarkur, north of Beersheva. There the fontwas situated in the westernmost of the seriesof rooms directly adjacent to the churchproper.24 On the other hand, the Phase Iarchitectural arrangement at Jabal Harûn wellreflects the requirement that the baptismalrooms should be attached to the church andprovided with direct comunication with it.25

The collapse of the extant (later in date)western wall of the chapel revealed the rubbleof the wall’s core and some ashlars whichprobably belonged either to the original

(early) wall of the chapel, or to the structureof the baptismal installation. Fragments ofpainted plaster were still attached to someashlars. In addition to some floral andgeometric designs, Greek letters or parts ofthem were also preserved. One such stonecontained a plaster fragment with thebeginning of the line, reading PRODROMO?[which can be interpreted as the epithet of Johnthe Baptist - prdro/moj “the Forerunner.”This epithet well corresponds to the baptismalfunction of this part of the chapel during PhaseI.

Phase II: Major Remodelling

It appears that a disaster, probably ofseismic nature, ended the Phase I occupationin the entire complex. The church was restoredbut also subdivided by a wall into the easternand the western part. The eastern part,internally ca 13 m (max) long, retained itsecclesiastical function but the columns wereremoved and replaced with free-standingpillars probably supporting E-W arches ratherthan architraves.26 Curiously, instead of apillar, the second support from the east in thesouthern row of support is a column. Later,this column received stone facing on all sides,resembling a cross. Throughout this phasenumerous changes and modifications tookplace in the bema area. The bema itself wasraised and laterally enclosed by two “counter-like” low walls, somewhat similar to those inPetra church in Phase V, or in the sanctuary ofthe monastic church at Deir ‘Ain Abata.27

Inside the apse, a thronos was inserted in themiddle of the synthronon. A marble floor waslaid out inside the apse. Only small fragmentsof the floor are preserved in situ, but themortar bedding impressions show that theslabs next to the synthronon were arranged ina curve parallel to the lower row of thesynthronon.

Inside the south pastophorion, a sandstonefloor was laid out. Close to the northern wall, a

43

large “tomb-like” underground space coveredby slabs was discovered but found empty. Thesize of that space (1.30 m long, 0.9 m wide,0.55 m high) appears small for a burial, but anossuary might be conceivable. East of the“tomb,” and directly in front of the cupboard,an equally engimatic installation wasconstructed. It included a rectangularenclosure made of thin sandstone slabs, with around hole (diameter 0.16 m) giving access toa sizeable pithos-like container under the floorlevel, and a stela-like construction made ofsandstone and marble fragments, set upright inthe middle of the enclosure. Given thepresence of the “tomb” next to it, the uprightinstallation might have been some kind of amemorial stone. The central feature appears tobe the pithos-container constructed of fiveseparate carved blocks of stone. The contentsof the container gave no indication what wasstored there (liquid?), but an ecclesiasticfunction seems possible. Probably during thisphase the interior of the pastophorion wasplastered over and a fragment from Psalm 90(91) was painted over in large Greek letters.This is probably one of the oldest,archaeologically recovered examples of OldTestament text.

The western part of the original church,ca. 9 m long, was apparently turned into anopen court (atrium) with two original E-Wrows of columns supplemented by the easternrow running N-S. No evidence for the westernrow of columns was detected so far, thus theatrium must have had two porticoes locatedopposite each other, and probably one on theeastern side. Initially, the old marble floor waspresumably in use. But later, that floor waspartially removed and replaced by the new(extant) floor which consisted of irregularsandstone slabs supplemented by brokenmarble pieces. This floor, laid out ca 0.2-0.25m above the level of the marble floor, ismarkedly sloping westward, as opposed to theoriginal floor. Apparently, heavy rains mighthave caused excessive flooding of the openatrium area. The new floor facilitated the

channelling of rainwater out of the atrium, andtoward the cistern.

During Phase I, some kind of an entranceporch might have preceded the church proper,being then followed by the courtyard with thecistern. In Phase II, a formal porch was erected- an enclosed space with a portico of fourcolumns in the front. The three channelswhich discharged rainwater from the atrium,run under the mosaic floor of the narthex.Originally, the mosaic featured an almostsymmetrical arrangement of designs on bothsides of the central door to the atrium,including armed humans and wild animals(Fig. 7). Except for the bordering chevronpattern, the geometric design in the center, andthe occasional fragments of human or animalbodies, the designs are not preserved since themosaic was heavily altered by latericonoclastic activities. Hunting scenes arewell represented in the mosaics of the 6thcentury, e.g., the mosaics at the HippolytusHall (6th century),28 or at the Old Diakonikon-Baptistery in the Memorial of Moses on Mt.Nebo (AD. 530).29 The central medallion ofthe Jabal Harûn mosaic - a complex geometricdesign of interlacing squares and ribbons - isstrikingly similar to the central panel of thenarthex mosaic in the church at Gharandal,tentatively dated to the 6th century.30

Major changes occurred also in thechapel. A new marble floor was laid out in theapse and in the area of a new transversal bemalocated in front of it. On the bema, a large altarmasonry base or pedestal was erected. Sincethe other half of this installation is still withinthe unexcavated northern balk, its dimensions(length ca. 0.88 m, width ca.1.4 m, height ca.0.78-0.97 m) are approximate. The structure ishollow inside, having a small compartment(0.54 m x 0.45 m x 0.65 m) with the openingtowards the apse. The marble fragment of aninscription, which reads ARVN, was found infront of the pedestal. The fragment could havebelonged to a marble sheating of theinstallation or to an edge of an altar tableplaced on top of it. The small compartment

44

inside the masonry pedestal might have servedas a depository of reliquaries which would beeasily accessible and available for display onvarious occasions. This would be generallyconsistent with the practices observed inPalaestina I, II and III and Arabia during theByzantine period.31 Although in this particularcase the reliquary would not be located in ashaft or fosse under the structure of the altaron the bema,32 but rather under the altar table,such cases are also known.33 Other reason forsuch location may be a relatively high level ofthe bedrock under the chapel and a shallowbuildup for the bema and the apse’s floors.The appearance of the fixed altar location inPhase II is also consistent with thechronological observations. The fixed altarappears relatively late in the 6th century, andonly at the end of that century special panelsor mosaic arrangements clearly marked theemplacement of the fixed altar’s supports.34

During the same phase, the baptismal fontin the chapel was abandoned and backfilled.The extant western wall of the chapel and abench against it were built then, the benchslightly overlapping the top of the font. Thewestern wall probably reduced the originallength of the chapel, and the new floor in thispart of the chapel completely covered theremaining part of the font. Probably, theinstallation of the masonry-built altar pedestalin the apse and the abandonment of thebaptismal font mark a significant redefining ofthe function of the chapel, perhaps in relationto the church or a chapel on the summit ofJabal Harûn, recorded by Wiegand (supra). Itis not possible to establish its constructiondate but equally nothing prevents this upperchurch to be considered coexistent with theearly monastery in Phase I. If the upper churchoriginally housed important relics, its possibledamage or destruction at the end of Phase Icould have caused the translation of the relicsdown to the rebuilt chapel of the monastery.

Fig. 7. The mosaic of the Jabal Harûn church.Photo: J. Mononen.

45

As such the chapel during Phase II would havebecome a memorial chapel and with that newfunction it might have been consideredunsuitable to retain the baptismal installationand practices there.

Phase III: Later Modifications

It is less certain whether Phase II was alsoended by a disaster but this remains a distinctpossibility. Resulting changes in the churchincluded the replacement of the pillars as roofsupport by pilasters and the N-S arches whichspanned the spaces of the nave and side aisles.Accordingly, the spaces between the free-standing pillars were walled up and pilastersbuilt against these walls and against the mainnorthern and southern walls of the church. Inthe southern aisle, two columns were usedinstead of the pilasters (Fig. 8). The bema areaseems to have been enclosed by a thick, poorly

built wall. Secondary walls built directly onthe pavements of the north pastophorion andin the area in front of the narthex may belongto this or later phases. Notably, such barriersand partitions are well-known from otherPalestinian churches in the Umayyad period.In the church of St. Mary at Rihab, and in theUpper Church at Quweisma, a secondary wallwhich connected the columns, effectivelyseparated the nave from the north aisle.35 Asimilar wall connected the pillars of the northrow in the church of St. John the Baptist (# 95)at Khirbet as-Samra.36

A massive buttress was built on theatrium’s side against the wall that partitionedthe early church in Phase II. The buttress builtas a wall-enclosing space filled with layers ofdebris, stones and reused material (includingcolumn drums) is currently ca 2.18 wide andca 2 m high. It stands directly on the upperfloor of the atrium.

It seems that during that period the

Fig. 8. Southern aisle of the church, looking west. The mixed column/pilaster supports of PhaseIII are visible on both sides of the paved area. Photo: Z.T. Fiema.

46

damage had been inflicted upon the mosaicfloor. The iconoclasts had removed not onlyalmost all tesserae forming faces but also mainparts of human and animal bodies, andreplaced them with plain large-size tesserae.The damaged areas appear as a result of acareful obliteration rather than a wantondestruction. Notably, this damage relates tothe images of animals and ordinary people, inoposition to the eighth century Byzantineiconoclasm that specifically targeted sacredimages. However, the perpetrators cannot beeasily identified since the destructionphenomenon is also present in some Jewishsynagogues in the region.37 At any rate, thiskind of deliberate damage which, neverthelesspreserves the mosaic in its entirety, isgenerally dated to the 8th century (lateUmayyad-early Abbasid period), and isknown from other churches in Jordan andPalestine. Particularly notable, deliberate butnot complete damage can be observed at theChurch of the Lions in Umm ar-Rasas.38 Theevidence of careful mosaic obliteration, as atJabal Harûn, should indicate that by the 8thcentury, the church would have been stillfunctioning in the ecclesiastical capacity.39

Later Phases

It is uncertain which parts of the entirestructure still retained their ecclesiasticalfunction during the later phases of thecomplex’s existence. Structural integrity ofthe building is also not supported as the apse’ssemidome seems to have collapsed by then ifnot earlier. Notable is the evidence of thecollection of marble fragments, glass andstone tesserae, and glass shards. Dumps orcollection points of such material have beenfound in the south pastophorion, in the ruinedapse of the church, in the cupboard next to thechapel’s apse, in various spots of the atrium,and in other places of the excavated parts ofthe complex. Some spaces within the churchand the atrium were temporarily or casually

occupied during later periods, a factexemplified by ashy spots, fireplaces and theabundance of bones (primarily, fish) in strataabove the original marble floor. Finally,substantial stone tumbles, including thecollapsed arches were documentedeverywhere. These episodes, either reflectingnatural decay and deterioration of structuralparts or subsequent seismic-relateddestructions, had definitely terminated theoccupation in the church area.

Historical Observations

Although the Byzantine monasticpresence at Jabal Harûn was already indicatedby the extant historical sources and earlyexplorations, the activities of the FJHPsubstantially contributed to a betterunderstanding of the function and thesignificance of the site. Although a directevidence is lacking, the association of themonastic presenceat Jabal Harûn, with apilgrimage center related to the veneration ofSt. Aaron appears reasonable. The lack ofliterary evidence specifically mentioningpilgrimages to the Mountain of Aaron may bedisturbing when comparing with theneighbouring Negev which certainly benefitedfrom the pilgrim traffic to Sinai.40 But thesanctity of the place continued to attract thepilgrimage traffic, whether Christian, Jewishor Muslim, also in the Medieval period,41 andit continued in the recent times. Therefore, theByzantine period should not be considered anexception. The presence of a large church withan associated chapel is certainly also animportant indicator of the nature ofoccupation. Notably, large, basilican churchesare rare in the monasteries of the JudaeanDesert.42 Additionally, the rooms around thenorthern wing of the complex appear to berelatively large; almost twice in size incomparison with known monks’ cells in someJudaean Desert coenobia.43 Most probably,these could be better interpreted as pilgrim

47

hostel. Furthermore, the discovery of theinstalations serving food producting andprocessing activities suggests that theMonastery of St. Aaron was a self-sufficienteconomic unit of a coenobion type.

Of great interest seems to be the SW areaof the complex which, at a certain point oftime became a disposal place. The largestdeposits in there were debris, includingbroken stones, ash, charcoal, mortar andplaster fragments and large quantities ofbroken ceramics. It seems that some stratamight have originated from the clearance aftera major fire which must have affected the site.Quite possibly, they may relate to thedestruction of the church in the early 6thcentury AD.

The ecclesiastical occupation in the JabalHarûn church appears to be much reduced inspace during later periods, yet some parts ofthe church seem to have continued in thatcapacity. But other, non-ecclesiastical (?)structures at the site, feature considerabledifferences in the intensity and character ofoccupation, often reduced to casual/temporarycharacter. A distinct feature, notable in almostall structures, including neglected orfunctionally modified parts of the church, isthe presence of concentrations of materialsuch as broken marble and glass fragmentsand glass and limestone tesserae. Someconcentrations may be interpreted as dumps ofdisused material but other as caches ofmaterial for remelting or burning for lime. Inthis respect, the later periods at the JabalHarûn complex reflect practices also noted onother Late Byzantine-Early Islamicecclesiastical sites in Jordan. Even moreelusive are the latest occupational periods forwhich the ecclesiastical occupation can nolonger be archaeologically confirmed. Forexample, it is impossible to establish a reliabledating of partitioning walls, simple enclosuresand campfires. But certain phenomena are ofparticular interest, such as the presence of RedSea parrot fish (Scaridae) among these lateremains, including the midden in the SW part

of the site. Notably, Scaridae appear commonin monastic contexts with a strong pilgrimagecharacter, e.g., at Dayr al-Qattar.44 Also, parrotfish appears in the earlier deposits at JabalHarûn. This should indicate the continuity ofthe pilgrims’ traffic to Jabal Harûn (and thecontinuous use of the midden), and thepreservation of traditional diet, even duringperiods when the monastic/ecclesiasticalstructures there were seemingly no longer inuse.

Although the post-Byzantine periods inthe history of the Jabal Harûn complex areparticularly enigmatic, some chronologicalindicators have been recovered. Both ceramicand glass material support the continuity ofoccupation throughout the 7th century AD,and the extension into the early 8th centuryappears to be attested by the presence ofcertain glass types, and by the specific type ofthe iconoclastic destruction of the narthexmosaic. Even the later 8th century may beconsidered through the presence of someceramic lamps, generally dated to the Abbasidperiod.45 Finally, the Crusader period, impliedin the written sources, seems to be nowconfirmed at the site by the presence of thestone scarp (glacis) on the western ridgebearing resemblance to the militaryarchitecture of the Crusaders, Ayyubids andMameluks.

Notes

1 I am thankful to Prof. Jaakko Frösén, FJHPDirector, and the entire FJHP team, includingthe Swiss ez-Zantur Project experts, for theircomments and advice. All errors of omissionand interpretation are mine.

2 See, R. Schick, “The Ecclesiastical History ofPetra,” forthcoming in the Petra ChurchProject, for the historical information andarchaeological data specifically concerning theChristians in Petra.

3 Palestinian grid coordinates of the shrine:188.64E x 969.667N; UTM coordinates731200E x 3356470N.

48

4 Eusebius, Das Onomastikon der biblischenOrtsnamen, E. Klostermann, ed. DieGriechischen Christlichen Schriftsteller derErsten Drei Jahrhunderte, vol. 11 (Leipzig,1904), 176, 177.

5 Leontios of Damascus, Vita di Santo StefanoSabaita. Bartolomeo Pirone, ed. and trans.(Jerusalem 1991), 16: 96-99

6 Abû al-Hasan ‘Alî al-Mas’ûdî, Kitâb al-Tanbîh wa-al-Ishrâf. ed. M. J. de Goeje(Leiden, 1894), 143-4.

7 Fulcher of Chartes, Fulcheri Carnotensishistoria Hierosolymitana (1095-1127), ed.Heinrich Hagenmeyer (Heidelberg, 1913), 2.5,381.

8 Gilbert the Abbott, "Gesta Dei Per Francos", inRecueil des Historiens des Croisades -Historiens Occidentaux, Tome IV (Paris,1879), 255. See also, "Gesta FrancorumExpugnatium Iherusalem", in Recueil desHistoriens des Croisades -HistoriensOccidentaux, Tome III (Paris, 1866), 523, and"Historiae Hiersolimitanae Secunda Pars", inRecueil des Historiens des Croisades -Historiens Occidentaux, Tome III (Paris,1866), 556.

9 Magistri Thetmari Iter in Terram Sanctamanno 1217, ed. T. Tobler (St. Gallen, 1851),30-33.

10 E.g., Ch. Irby and J. Mangles, Travels in Egyptand Nubia, Syria and Asia Minor (London,1823) 433-9; E. H. Palmer, "The Desert of theTîh and the Country of Moab", PEF QuarterlyStatement (1871), 50-1.

11 A. Musil, Arabia Petraea II. Edom 1. (Wien,1907), 113-8, 161; Th. Wiegand. Sinai.Wissenschaftlische Veröffentlichungen desDeutsch-Türkischen Denkmalschutz-Kommandos, Heft 1 (Berlin, 1920), 136-45.

12 G. L. Peterman and R. Schick, "The Monasteryof Saint Aaron", ADAJ 40 (1996), 473-480.

13 The documents are studied by the team fromthe University of Helsinki (Finland), whichalso undertook the conservation of the papyri,and from the team from the University ofMichigan, headed respectively by Profs.Jaakko Frösén and Ludwig Koenen. The firstvolume of the publication series has justappeared: J. Frösén, A. Arjava, and M.Lehtinen, The Petra Papyri I, (Amman, 2002).

14 J. Frösén and Z. T. Fiema, "The Petra Papyri",ACOR Newsletter 6.2 (1994), 1-3; T. Gagos

and J. Frösén, Petra Papyri, ADAJ (1998),476-77, inv. 6a; 480, inv. 86.

15 For the results, see J. Frösén et al., "TheFinnish Jabal Haroun Project Report on the1997 Season", ADAJ 42 (1998), 483-502; J.Frösén et al., "The 1998 Finnish Jabal HarûnProject. A Preliminary Report", ADAJ 43(1999), 369-410; J. Frösén et al., "The 1999Finnish Jabal Harûn Project: A PreliminaryReport", ADAJ 44 (2000), 395-424; J. Frösénet al., "The 2000 Finnish Jabal Harûn Project:Preliminary Report", ADAJ 45 (2001), 359-376, and J. Frösén et al., "The 1988-2000Finnish Jabal Harûn Project: SpecializedReports", ADAJ 45 (2001), 377-92.

16 Z. T. Fiema, “Reconstructing Culture History ofthe Petra Church: Data and Phasing,” in Z. T.Fiema, Ch. Kanellopoulos, and T.Waliszewski, "The Petra Church", (Amman,2001).

17 J. W. Crowfoot, Early Churches in Palestine,(London. 1941), 54, 61; R. H. Smith, and L. P.Day, Pella of the Decapolis. Vol 2. (Wooster,1989), 84. See also A. Negev, "The Churchesof the Central Negev - An ArchaeologicalSurvey", RB 81 (1974) 400-11. Themonoapsidal type in Syria is generally dated tothe fifth century, see H.C. Butler, EarlyChurches in Syria. Fourth to SeventhCenturies. Part One. History, ed. E. BaldwinSmith (Princeton, 1929), 48-82

18 Generally, the type of bema does not constituteany significant chronological marker (R.Rosenthal-Heginbottom, Die Kirchen vonSobota und die Dreiapsidenkirchen des NahenOstens. Wiesbaden, 1982, 149, 151).

19 M. Ben-Pechat, "The Paleochristian BaptismalFonts in the Holy Land: Formal and FunctionalStudy," LA XXXIX (1989), 173-4; "Baptismand Monasticism in the Holy Land:Archaeological and Literary Evidence,” inChristian Archaeology in the Holy Land. NewDiscoveries," eds. G. C. Bottini, L. di Segni,and E. Alliata (Jerusalem, 1990), 510; M.Piccirillo, "I Battisteri Bizantini di Giordania,"in Noscere Sancta Miscellanea in Memoria diAgostino Amore OFM (+1982). Volumeprimo: Storia della Chiesa, Archeologia, Arte.I, ed. V. Janeiro (Roma, 1985), 355.

20 Ben-Pechat, "The Paleochristian," fig 1;Rosenthal-Heginbottom, Die Kirchen, 174-200 (baptisteries in the Negev).

49

21 Ben-Pechat, "Baptism," 501-2.22 B. Bagatti, The Church from the Gentiles in

Palestine. History and Archaeology(Jerusalem, 1984), 305; Ben-Pechat,"Baptism," 508-9.

23 M. Piccirillo, "Campagna Archeologica nellaBasilica di Mose Profeta sul Monte NeboSiyagha," LA XXVI (1976), 305-12; A.Michel, "The Liturgical Installations," inMount Nebo. New Archaeological Excavations1967-1997, eds. M. Piccirillo and E. Alliata(Jerusalem, 1998), 405.

24 P. Figueras, "L’église byzantine de Karkourdans le Néguev," in Acta XIII CongressusInternationalis Archaeologiae Christianae III(Città del Vaticano - Split, 1998), 263-4, Fig.2.

25 Ben-Pechat, "Baptism," 508-9.26 Notably, the same roof-supporting

arrangement is known from the Lower Church(C101) at Humeima, see R. Schick,"Christianity at Humayma, Jordan," LA XLV(1995), pp. 324-5.

27 For Petra, see Fiema, "Reconstructing;" forDeir ‘Ain Abata, see, K. D. Politis, "The 1992Season of Excavations and the 1993 Season ofRestorations at Deir ‘Ain ‘Abata," ADAJ 37(1993), 507, fig. 6.

28 M. Piccirillo, The Mosaics of Jordan(Amman, 1993), 23-24; 58-59.

29 Ibid., 135, 14630 N. Ricklefs, "The Church Mosaics," in A.

Walmsley, "The Church at Arindela(Gharandal) of Palaestina Tertia," LiberAnnuus XLVII (1997), 501-3, fig.5.

31 P. Donceel-Voute, "Le role des reliquiairiesdans les pèlerinages," in Akten des XII.Internationalen Kongresses für ChristlicheArchäologie (Münster, 1995), 198-200; "Lamise en scène de la liturgie au Proche OrientIVe-IXe s.: les 'provinces liturgiques'," in TheChristian East, Its Institutions & Its Thought.ed. R. F. Taft (Roma, 1996), 328.

32 As at Umm al Rasas, see E. Alliata, "IReliquiari e Altri Elementi Architettonici," inUmm al-Rasas—Mayfa’ah I. Gli Scavi delComplesso di Santo Stefano, eds. M. Piccirillo

and E. Alliata (Jerusalem, 1994), 312-14; andA. Michel, "Le Installazioni Liturgiche," inUmm al-Rasas—Mayfa’ah, I, eds. M. Piccirilloand E. Alliata (Jerusalem, 1994), 117, 119,note 15.

33 Michel, "The Liturgical," 394.34 N. Duval, "L’architecture chrétienne et les

pratiques liturgiques en Jordanie en rapportavec la Palestine," in Churches Built in AncientTimes. Recent Studies in Early ChristianArchaeology, ed. K. Painter (London, 1994),170, 203.

35 M. Piccirillo, "The Umayyad Churches ofJordan," ADAJ 28 (1984), 338.

36 J-P. Humbert and A. Desreumaux, "Huitcampagnes de fouilles au Khirbat es-Samra(1981-1989)," Revue biblique 97/2 (1990),261. See also partitioning walls built in thenave and the aisles at the Anchor Church (theAbbasid phase) at Tiberias, which effectivelydivided the interior into several rooms orcompartments (Y. Hirschfeld, "The AnchorChurch at the Summit of Mt. Berenice,Tiberias," Biblical Archaeologist 57/3, 1994,126, 132).

37 Piccirillo, The Mosaics, 42.38 Ibid., 211.39 Ibid., 42.40 P. Figueras, "Pilgrims to Sinai in the Byzantine

Negev," in Akten des XII. InternationalenKongresses für Christliche Archäologie,(Münster, 1995), 756-762.

41 As attested by Hebrew medieval inscriptionsleft on the stone sarcophagus inside the weli.

42 E.g., Y. Hirschfeld, "List of the ByzantineMonasteries in the Judaean Desert," inChristian Archaeology in the Holy Land. NewDiscoveries, eds. G. Claudio Bottini, L. DiSegni and E. Alliata, (Jerusalem, 1990), 63-65,Site 57; The Judaean Desert, 130.

43 Hirschfeld, The Judaean Desert, 176-7.44 R. Holmgren and A. Kaliff, "The 1995-1996

Excavation of Dayr al-Qattar al-Byzanti:Preliminary Report," ADAJ 41 (1997), 324.

45 See, A. Karivieri in Frösén et al. "The 1998,"389-91. See specialized pottery and glassstudies in ADAJ 45, 2001.