Northern Jordan: What Might Have Been in the Madaba Mosaic Map

by Robert Schick

The East Edge of the Map

What Certainly Would Have Been Depicted

What Might Have Been Depicted

The Onomasticon of Eusebius

Other Sources for Placenames

The Madaba Region

This paper presents the information about northern Jordan available in geographical sources of the Roman-Byzantine period in order to suggest what might have been represented in the now missing northeast portion of the Madaba mosaic map. It will cover the area from north of Karak to the Yarmouk River along the Jordanian-Syrian border.

The East Edge of the Map
But first, I wish to address the question of the original eastern edge of the map in relation to the easternmost piers. The reconstructions of the church plan show the easternmost pier in the south row and a chancel screen marking the eastern limit of the map. The precise location of the chancel screen is described by Manfredi (1899: 151), who reported that the southwestern pilaster (into which is cut a groove for insertion of a marble panel) of the chancel screen was flush with the north face of the easternmost pier in the south row and touched the easternmost point of the map. This is shown on some of the reconstructions (e.g. Clermont-Ganneau 1987: 214). The published reconstruction of the Byzantine church in Avi-Yonah's study (1954: 14, fig. 3) again has the screen lining up with the center of the easternmost pier, but has the easternmost preserved part of the map - the portion east of Karak - extending east of the west edge of that easternmost pier, so that the map comes very close to touching the marble panels themselves. Avi-Yonah (1954:15) suggests the distance as only 20 cm.
But Avi-Yonah's reconstruction fails to take into account the width of the grooved blocks into which the screen was set, plus the additional rows of cubes needed along the east edge of the map to complete the upper part of the city vignette east of Karak. Avi-Yonah's reconstruction also seems to assume that the area east of the screen was at the same level as the aisles. If one maintains that the chancel area was elevated (as does Piccirillo 1989: 81, for example, based on Manfredi 1899), then the width of the one or two steps up west of the chancel screen panels also would need to be factored in while estimating the distance between the preserved map and the chancel screen panels. That alternative of a raised chancel would throw Avi-Yonah's reconstruction off even more.

Northern Jordan in the Surviving Portion of the Mosaic Map

Three places are identified by name in the surviving portion:
1) Aenon, now Sapsaphas, marking the spot where John baptized, repeating the Aenon that occurs further north on the map
2) Baaru, the hot springs of Zarqa Ma'in
3) Hot Baths of Callirhoe, the hot springs of 'Ayn Sara.

Small portions of two other unlabelled city vignettes survive. A variety of suggestions have been made for their identification. The city vignette above the gazelle to the left could be Abel (Khirbat al-Kafrein) or Livias/Beth-Ramtha (Tell al-Rame), or perhaps Beth Nimrin (Tell Nimrin), while the city vignette above the gazelle to the right could be Beth Jeshimoth (Suweima).

What Certainly Would Have Been Depicted
To turn to what would have been depicted on the map, the mosaicists would surely have indicated the tribes of Israel located east of the Jordan River. That would be: the "Lot [Klhros] of Reuben" in the area around Madaba and the "Lot of Gad" in the area west of 'Amman. The label "The Lot of Manasseh" could have been put either in the area of north of 'Ajlun or west of the Jordan, north of Nablus.
More of the Wadi Mujib, Wadi Zarqa Ma'in and the Zarqa River would have been depicted, along with more hills in the Karak area. Another prominent geographical feature that is a candidate for being depicted on the map is the Yarmouk River, while northern Jordan is hilly enough for the mosaicist to have chosen to represent some hills.

What Might Have Been Depicted
As for the localities that might have been depicted on the map, there are a large number of place names recorded in the geographical sources to choose from - more than could reasonable have been depicted on the map. Basic information about the place names recorded from the Roman-Byzantine sources can be found conveniently in Avi-Yonah's Gazetteer of Roman Palestine (1976); Abel's Géographie de la Palestine (1938) is another standard source. Avi-Yonah lists some 60-odd place names for the area under consideration.
The available space on the map appears large enough for a couple dozen city vignettes and other identifications, meaning that perhaps as much as a third to a half of all the attested place names could have been depicted.

The Onomasticon of Eusebius
Thirty five places in the area are listed in Onomasticon of Eusebius. A list of those placenames follows, along with Avi-Yonah's suggested identifications, which in some cases remain unconfirmed, and the page and line reference of Eusebius.

North of Karak and south of the Wadi Mujib are four sites:

Dannaea (Khirbat al-Dann) - 76.10
Eglaim (Rujm al-Jalame) - 36.20
Luhith (Khirbat al-Dubab) - 122.29
Rabbath Moab (Rabba) - 10.17, 36.24, 124.17

North of the Wadi Mujib and south of the Zarqa River are 23 sites:

Abila 2 (Khirbat al-Kafrein) - 32.15
Aroer ('Ara'ir) - 12.5
Baaras (Zarqa Ma'in) - 44.22, 112.17 - depicted on the map
Beelmaus (Ma'in) - 46.1
Beth Jeshimoth (Suweimeh) - 48.7
Beth Nimrin (Tell Nimrin) - 44.17
Beth Peor (Khirbat al-Shaykh Jayil) - 48.3
Bethany-Bethabara (Tell Medes) - 58.18
Cariatha (al-Qurayat) - 112.15
Dibon (Dhiban) - 76.18
Elealeh (al-'Al) - 84.10
Enganna ('Ayn Jinna) - 94.26
Heshbon (Hesban) - 84.4
Jazer (Khirbat al-Sar) - 104.13
Maanith (Umm al-Hanafish) - 132.2
Machanarath (al-Manakir) - 126.18
Maschana (al-Madina) - 126.14
Medaba (Madaba) - 128.19
Mefa (Umm al-Rasas) - 128.21
Nebo (Khirbat al-Mukhayyat) - 136.6, 136.13
Philadelphia ('Amman) - 16.15 12.1, 24.2
Ramoth (Khirbat Jal'ad) - 144.5
Zia (Zay) - 94.3

North of the Zarqa River and south of the Yarmouk River are eight sites:

Abila 1 (Abila) - 32.16
Amathus (Tell 'Amtah) - 22.24
Arbela (Irbid) - 14.19
Erga ('Arjan) - 16.21
Gadara (Umm Qays) - 74.10
Gerasa (Jerash) - 64.2
Jabesh Gilead (Tell al-Ras) - 110.12
Pella (Tabaqat Fihl) - 14.19, 22.25, 32.6, 80.17, 110.13

Succoth (Deir 'Alla) in the Jordan Valley, with its multiple Old Testament associations, is oddly absent from the Onomasticon.

Machaerus (Makawar) is also absent from the Onomasticon; the tradition that it is where John the Baptist was killed derives from Josephus, not the Gospel accounts.

Other Sources for Placenames
One indication that the Onomasticon of Eusebius was not the only source for the places depicted on the Madaba Mosaic Map is the fact that two of the three places in northern Jordan that are depicted on the surviving portion of the map are not listed in the Onomasticon: "Aenon, now Sapsaphas" and "Hot Baths of Callirhoe". Although Baarou, the third site, is both on the map and listed in the Onomasticon, clearly the mosaicist did not restrict himself to selecting sites only from the Onomasticon.
Another principal of selection of sites for the map may have been the location of sites on the network of roads. For northern Jordan, the routes depicted on the Peutinger Table record several places not listed in the Onomasticon: Thantia-Thainatha (Thughrat al-Jubb, rather than Umm al-Jimal; see Kennedy 1982: 152-154), Hatita (Khirbat al-Samra); Gadda (al-Hadid), and Capitolias (Bayt Ras), while Philadelphia (Amman) and Gadara (Umm Qays) are also listed in the Onomasticon. Absent from the Peutinger Table are places on the Via Nova south of Amman, such as Hesban, Madaba, or Dhiban, while Jarash is also conspicuous by its absence.
Other placenames, such as the ones known only from attestations in such sources as the writings of Polybius, or the fifth century A.D. Notitia Dignitatum list of military garrisons are less likely candidates for consideration for inclusion on the Madaba mosaic map.

The Madaba Region
The map is laid out so that Jerusalem is at the center. By happy coincidence Madaba itself is on the same central axis farther to the east. And so the home city of the map would have had a prominent location directly in front of the entrance to the chancel area. One might also think that local pride would have ensured the inclusion of Umm al-Rasas, Ma'in, and Mount Nebo, perhaps along with Libb, all recorded in the eigth century mosaic floor in the Church of Saint Stephen at Umm al-Rasas.


Abel, F.-M.
1938 Géographie de la Palestine. Paris: Lecoffre.

Avi-Yonah, M.
1976 Gazatteer of Roman Palestine. Qedem 5. Jerusalem: Hebrew University.

Clermont-Ganneau, Ch.
1897 The Madeba Mosaic. Palestine Exploration Fund Quarterly Statement: 213-225.

Kennedy, D.
1982 Archaeological Explorations on the Roman Frontier in North-East Jordan. BAR International Series 134. Oxford: British Archaeological Reports.

MacAdams, H.
1994 Settlements and Settlement Patterns in Northern and Central Transjordania, ca 500-ca 750. Pp. 17-93 in G.R.D. King and Averil Cameron, eds., The Byzantine and Early Islamic Near East II: Land Use and Settlement Patterns. Princeton: Darwin.

Manfredi, G.
1899 Piano generale delle antichità di Madeba. Nouvo Bullettino di Archeologia Cristiana 5: 149-170.

Piccirillo, M.
1989 Chieses e Mosaici di Madaba. Jerusalem: Franciscan Printing Press.

This contribution was first published in: The Madaba Map Centenary, Jerusalem 1999, 227-229.

logo logo

Created Saturday, December 16, 2000 at 11:31:34
by Eugenio Alliata ofm in collaboration with Stefano de Luca ofm
Webmaster: John Abela ofm - Space by courtesy of Christus Rex
copyright - Studium Biblicum Franciscanum - Jerusalem 2000