The Uniqueness of the Madaba Map and its Restoration in 1965

by Herbert Donner

I. [Uniqueness]

II. [1965 Restoration]

We might expect that the Madaba Mosaic Map should play an outstanding part in the history of oriental and occidental cartography. Strangely enough, it seems to be not so. The Madaba map is not even mentioned in the majority of the great encyclopedias and lexica such as the Encyclopaedia Britannica, the Colliers Encyclopedia, the Brockhaus Enzyklopädie in 20 volumes and others, or at any rate not under the key words cartography, geography and itinerary. It is also missing in the fine and extensive book about the early history of cartography by written by Grosjean and Kinauer (1975), although the authors devoted no less than five pages to the Peutinger Tables. It is difficult to understand why that is so. For the special importance of this incomparable monument has been emphasized more than once by several scholars from different countries (e.g. Jacoby 1905: 28; Avi-Yonah 1954: 9; Donner 1992: 19 ff.). Let us hope that this conference will contribute to eliminate that state of underestimating the Madaba map.

I do not want to anticipate the numerous papers of this conference, but only to make some remarks about the uniqueness of the Madaba map in the history of ancient cartography and mosaic art. It seems advisable to begin with a simple definition: a monument may be called unique if it has no comparable parallels from its own time and world. Immediately, the objection will be raised: then the Madaba map cannot be regarded as unique, for there are parallels, parallels of different sort, e.g. the small Palestine section of the Peutinger Plates, the mosaic representations of cities known from Ma'in, Umm al-Rasas, and also from Ravenna, Rome and from elsewhere, not to forget the Roman-Byzantine landscape mosaics from the Africa proconsularis, from Italy (especially Pompeii), indeed from almost all countries of the Roman oikumene, even including at-Tabgha. Nobody should deny, of course, that there are relations between these monuments and the Madaba map. The geographical and artistic traditions which produced them may have had an influence on the Madaba map, and it is worth thoroughly investigating such details. But all these monuments are no real parallels to the Madaba map. The Peutinger Plates, to begin with, constitute a medieval copy of a road map of the whole Roman-Byzantine Empire, probably originating in the fourth or even fifth century A.D. (see below the contribution by Ekkehard Weber) and either revised by Christians or of Christian origin in general. Although the mosaic artist of Madaba may have used Byzantine road maps as sources, the mosaic map itself is obviously by no means a road map. We may ask of a road map the actual depiction of roads such as in the Peutinger Plates, which are full of roads. On the Madaba map only one road is really represented in white cubes: the first section of the junction road between Jerusalem and Neapolis, starting from the Damascus Gate and running between the lines three and four of the blessing of Benjamin according to Deut 33:12. Other roads are indicated but not depicted: from Jerusalem over Nicopolis ('Amwas) to the coastal plain by the mile-stations TO TETARTON "the ninth (milestone)", from Jericho to Scythopolis (Beisan) by the sites ARCHELAIS, (FASAE)LIS, KOREOUS and SA(LOUMIAS), the via maris between Gaza and Pelusium (Tall Farama) by no less than eight sites, and - if I am not mistaken - the three main roads running approximately along the collateral lines of the Delta triangle in Lower Egypt, namely from Pelusium to Memphis, from Alexandria to Memphis, and from Pelusium to Alexandria. Here we have indications only, and how many important roads are completely lacking, neither being represented nor indicated, will become clear by comparison with a modern historical map of Roman-Byzantine Palestine.
As far as the representations of cities or villages are concerned, the mosaicist of Madaba used symbols resembling those of Hellenistic-Roman origin, widespread in all Mediterranean countries: from the simplest shape with two towers and one gate in between to larger and more differentiated representations, even to peculiar and highly imaginative pictures as e.g. in Lower Egypt, made in quite another style compared to the Palestinian cities. Such symbols do not show any of the individual characteristics of the represented cities or villages. For some larger cities, however, the mosaicist of Madaba preferred a kind of oblique pictorial representation, totally unknown from other mosaics, although still depending on the Hellenistic-Roman tradition for depicting cities. The distinguishing feature is exactness. What we have here appears to be an early form of a city-map. Those representations are of Jerusalem, Neapolis, Ascalon, Gaza, Pelusium, Charachmoba, and up to a certain point Eleutheropolis, Diospolis and Jamnia as well. Some of these cities are depicted so precisely that we can orient ourselves by the representations up to present, such as Jerusalem or Gaza, others show the old city centers still preserved in the last century, such as Neapolis and al-Karak, and some of them were uncovered by archaeological excavations in our century, such as Ascalon and partly Pelusium. There are absolutely no parallels for this kind of representation; the Madaba map is in this respect, as far as I see, quite original and unique.
Finally, the fine and beautiful landscape mosaics, widely attested since Hellenistic times, cannot seriously be taken into consideration as parallels to the Madaba map. For they do not intend to portray facts and features of real geography and topography; they rather are made to artistically produce an atmosphere or a kind of mood wanted by the customers or donors. This is clearly recognizable from the so-called Nile mosaics illustrating what we call Egypt romanticism of late antiquity. The mosaic map of Madaba, however, is not a landscape mosaic, but a real geographical map apparently intending to depict real geographical and topographical facts. In this respect it is totally unique in the whole of antiquity.
The mosaic artist of Madaba tried to represent as precisely as possible the nature of Palestine on both sides of the Jordan River and to establish the locations of Palestinian sites. Naturally, he did not succeed in that everywhere, but still he did to a considerable and astonishing degree. The main difficulty consisted in the fact that the art of accurate scale map-making was not yet invented in the sixth century A.D. Nobody in antiquity seems to have had the idea to use the square system of the Roman land surveyors (agrimensores) for the cartographic representation of regions or even of whole countries. Nevertheless, the cartographic accuracy of the mosaicist is significant: of course not an absolute, but a relative accuracy. Look, for example, at the characteristic dislocation of the mountain range east of the Dead Sea, forming three mountain levels, or at the change of direction of the Wadi Zarqa Ma'in from north-south to east-west, or at the mouth of the Arnon River (Wadi al-Mujib) north of the middle of the Dead Sea, or at the correct differentiation of mountainous and plain areas, etc. The mosaicist had special difficulties in relating the places to each other correctly. He did not succeed in solving this problem everywhere, but more frequently than one might expected. For example, Jerusalem, or at least its northern wall, is exactly situated at the latitude of the northern end of the Dead See, Hebron (al-Khalil) approximately at the latitude of Eleutheropolis (Beit Jibrin), Livias correctly east-southeast of Jericho, Mampsis (Kurnub) east-southeast of Beersheba (in reality southeast), Elusa and Orda southeast of Gaza in spite of the wrong turning of the Mediterranean coast line, etc. Very interesting is the representation of the plain of Shechem/Nablus: this plain extends from west to east, but is turned up to the south on the Madaba map, the symbols being in correct relationship to each other, namely Sychar, Sichem, Joseph's tomb and Jacob's well.

To explain the geographical and topographical inaccuracies and mistakes that occurred (for the main ones see Donner 1992: 18), we may point to four issues: 1. the lack of a precise scale for the map in all its parts; 2. the lack of space, caused not only but mainly by the excessive size of some city depictions, such as Jerusalem and Neapolis; 3. the "religious geography", i.e. the overriding importance of religious convictions to real geography; 4. the lack of familiarity with some parts of the land (e.g. the area southeast of the Dead Sea and in Lower Egypt). The errors or inaccuracies cannot diminish the high cartographic value of the map. Even with regard to its cartographic exactness the map is without any real parallels up to now, therefore it may be called unique.
Of course, we speak of uniqueness only with the reservation that no other real comparable maps of this kind will come to light. This has not happened until now. We have to leave it open whether it will happen in the future or not. Considering the numerous mosaics uncovered in Jordan during our century, it becomes more and more improbable. In Madaba itself, however, other mosaic maps of the same style, from the same school and even by the same artists might be found. Moreover, a lot of mosaics seems to have been destroyed since Madaba was rebuilt around 1880. Perhaps, the so-called fragment C, representing the Phoenician Sarepta, published by J. Germer-Durand (1895: 588) belongs here; it is said to have been in the corridor of a private house at Madaba, but it has totally disappeared. Considering the extension of the mosaic map to the north and its position in the present church, fragment B with the blessing of Zabulon near the northern wall of the modern church and of the ancient basilica, it seems rather improbable that fragment C ever belonged to the Madaba map. But this has to remain open, for there are strong reasons to assume it could nonetheless have belonged to it (see below the contribution by Patricia Bikai). At any rate, we might draw the conclusion that mosaics similar to the map existed elsewhere in Madaba. Musil's brief communication (1907: 116) concerning the Madaba mosaics is enigmatic and not easily understandable. They run as follows (translated from German): "On the left, immediately behind the gate, there were still in 1896 entire areas covered with mosaics showing all sorts of scenes: a man, e.g., killing a wild ass with a spear, antelopes, fleeing gazelles, human beings in resting or dancing position and so on. East of these mosaics, there was at that time the atrium or the portico of a basilica, the foundations walls of which were still clearly visible. There was a mosaic floor as well, and the catholic missionary Father Biever (actually in Tabgha at the Lake of Gennesaret) told me in 1898 that he saw in this place in the eighties a mosaic map and copied some parts of it. It was meant to be represented on this map even Babylon and Rome." Babylon and Rome? Certainly not on our mosaic map, whose extent we can reconstruct at least approximately. Shall we assume that Father Biever was not able to read Greek letters sufficiently? Or did he see another, no longer extant mosaic, perhaps in front of the facade, i.e. west of the Church of St. George or elsewhere? We do not rightly know what to make of it.
The idea of uniqueness of the Madaba map could be pursued in other directions, but we would have to leave the fields of geography, topography and cartography, and enter theology. The exceptional idea, totally unknown before the sixth century, to illustrate God's salvation history in a map, was formed and realized only once, as far as we know: at Madaba. I do not intend to investigate that further. Let us confine ourselves to call out to the historians of cartography: Turn your eyes inward, and allocate the mosaic map of Madaba the place due to it, the place at the beginning of the history of cartography, because it is the oldest known map that deserves this name!

Around the middle of our century, the Madaba map was in such a poor condition that careful restoration and conservation was urgently needed, in order to prevent its irrevocable destruction. It is not that the builders of the present church treated the mosaic floor improperly, as has customarily been assumed since Musil (1907: 116). On the contrary, we may certify that the architect and the workers took utmost care over it; they were by no means without any sense of responsibility. The main reason why the condition had become so bad was a double one: on the one hand the very long and very heavy wooden cover planks which began to swing when handled by only one person, and which often crashed upon the mosaic floor, and on the other hand the vast quantities of water poured all the time upon the mosaic in order to make its colors bright. People did not take into account that water is poison for ancient mosaics. From that, the mosaic had peeled off from its original lime bed here and there, and had bulged out so that one could move it down by thumb 1 or 2 cm. Parts of the mosaic, especially in the region of Jerusalem, were held by surface tension only; they were in permanent danger of breaking into pieces by the impact of the planks or by someone walking on it.
The restoration and conservation of the Madaba map was more than 30 years ago. The details were reported in 1967 (Donner and Cüppers 1967: esp. 3-13). I shall confine myself to the essential facts. In December 1964, the Volkswagenwerk Foundation in Hannover generously granted the German Palestine Exploration Society DM 90,000, a rather large amount at that time. We succeeded in getting two excellent mosaic experts: the former custos and present director of the Rheinisch Museum at Trier, Dr. Heinz Cüppers, and the museum's late restorator, Mr. Heinrich Brandt.
After we made a preparatory trip in April 1965, and obtained the necessary ecclesiastical and state permits, we started on September 15 1965 and finished on November 14, 1965 with the ceremonious unveiling of the mosaic in the presence of His Beatitude Benedictos I, the late Patriarch of Jerusalem and All Palestine, and numerous political and ecclesiastical dignitaries.
After having carefully cleaned the surface and the edges of the mosaic and removed the cement filling in the damaged spots, we divided the mosaic into 14 segments, making the cuts in such a way that they followed the larger outlines (rivers, shore of the Dead Sea, mountains) and avoided the smaller representations and inscriptions. After that, we spread a glue, called Mowilith, on the whole surface and applied a tightly adhering cotton on top. The segments were carefully separated from each other along the cut lines and turned over. Now we could see numerous mosaic cubes, originally up to 10 cm long, broken in the middle or more to the top. This forced us to remove the whole ancient lime bed so that the cubes or what had remained of them would adhere to the cotton alone. Thus, each segment looked like the skin of a hedgehog with spines of different length. Then we made a new bed, approximately 4.5 cm thick, consisting of a mixture of concrete and sand, wire netting and another concrete layer (seen from above). The subsoil was dug up, half a meter deep. Nowhere did we find clearly recognizable remains of a preparatory sketch. We laid (now seen from below) a limestone layer of 15-25 cm, above it a concrete slab of 25-35 cm covered with a thin asphalt layer in order to prevent the rise of ground humidity into the mosaic. After having put back the segments on the floor, we realized that the mosaic had grown a little broader because of the new bed and of the new horizontal base, mostly by 1 cm at the edges, but at the Nile Delta, however, up to 4.5 cm, because the mosaic had sloped down towards the south wall of the church. The separating cuts between the segments had also become a little wider here and there; in such cases we needed to supplement one or two rows of white mosaic cubes not belonging to the original surface of the map but taken from the subsoil. Dr. Cüppers reported on the details. The restored mosaic was cleaned once more and was finally painted with a transparent varnish in order to preserve the brightness of the colors, but the varnish performed its task only a few years.
Since then, I have rather often been in Madaba, sometimes with fear and trembling, in which I even caused H. R. H. the Crown Prince Hassan to participate, I think it was at the First Conference on History and Archaeology of Jordan at Oxford. For colleagues of mine and tourists told me in Germany that the condition of the map had worsened again in the meantime. But I could see on each visit, last time in October 1995, that this was fortunately wrong. The mosaic is definitely fixed in its new bed, and no cube is movable - only that is of decisive importance. The impression of worsening may have been caused by the amount of dirt covering the surface again, perhaps as well by the disappointment at the somewhat dull colors when no water is poured upon the mosaic - as it seems to be, I will hope at least. At the end of this conference we shall discuss the future preservation of the mosaic map. It appears to me that it would be sufficient to clean the surface from time to time. A renewal of the transparent varnish that has disappeared long ago could be useful, not so much for the mosaic itself, but for the visitors. I do not rightly know whether the carpeting, which we had put on the mosaic to protect it and which also disappeared, should be renewed; probably, it was more trouble than good.
In any case, there is good reason to assume that this unique monument will enter the third millennium A.D. and the sixteenth century after the Hijra without further damage.


Avi-Yonah, M.
1954 The Madaba Mosaic Map with Introduction and Commentary. Jerusalem.

Donner, H. and Cüppers, H.
1967 Die Restauration und Konservierung der Mosaikkarte von Madeba. Zeitschrift des Deutschen Palästina-Vereins 83: 1-33.

Donner, H.
1992 The Mosaic Map of Madaba. An Introductory Guide. Palaestina antiqua 7. The Netherlands.

Germer-Durand, J.
1895 Inscriptions romaines et byzantines de Palestine.Revue Biblique 4: 587-592.

Grosjean, G. and R. Kinauer
1975 Kartenkunst und Kartentechnik vom Altertum bis zum Barock, 2nd ed. Bern and Stuttgart.

Jacoby, A.
1905 Das geographische Mosaik von Madeba. Studien über christliche Denkmäler, Neue Folge 3. Leipzig.

Musil, A.
1907 Arabia Petraea. I. Moab. Topographischer Reisebericht. Vienna.

This article was first published in: The Madaba Map Centenary 1897-1997, Jerusalem 1999, 37-40.

logo logo

Created Saturday, December 16, 2000 at 11:13:51
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