The Roman Temples of Lebanon by George Taylor
Copyright Dar el Machreq Publishers, Beirut

Introduction

Of all the diverse architectural achievements of the Greeks and Romans, the temples built for the immortal gods best illustrate Classical perfection in the fine arts. This is not surprising – indeed, it would be strange if it were not so – since the search to give form to a divine image, and the need to erect a shelter worthy of that image, have been the motivating force for a number of the fine arts, particularly sculpture.

The gods were worshipped, of course, long before temples were built to house their images. The high mountains, the sources of rivers and streams, caves, and trees, were all thought of as the homes of the gods, and in these natural surroundings – unadorned by man – the gods were first worshipped (Pl. 101). The coolness of mountain and spring, the dim light of a cavern, the shade of a tree, must have been very welcome in the fiery Mediterranean sun, and it seems not unnatural that such places were chosen as the homes of the gods. Of course, coolness and shade did not alone determine the place where a god was to be worshipped; the site had perforce to arouse sacred awe in the mind of the worshipper. The offering of propitiatory gifts, and the need for a receptacle to hold them, would have led to the use of altars at these sacred sites; and, in time, the need to protect the altars must have led to the first temple buildings.

It seems reasonable to relate the adoption of a houselike shelter for the god to the adoption of a human representation of the god: the more human the god’s image became, the more closely his “house” resembled the human dwelling house. And, since a god worthy of veneration had to be honoured by a dwelling house worthy of admiration, it is likely that the most refined of human dwelling houses, the royal palace, was chosen as the model for the early temple

It is tempting to see in the plan of a Homeric palace the essential ground plan of the Greek and Roman temple: the outer wall, the courtyard, the megaron of the palace corresponding to the enclosure wall, the temenos, the sanctuary of the temple. This is a gross over-simplification of the development of the temple, of course, but it is a useful comparison to have in mind when considering the Roman temples in Lebanon. A massive wall encloses Lebanese temples where the site permits the leveling of a large terrace; there is a paved area between the wall and the sanctuary; and the sanctuary itself is raised – often on immense substructures – to a commanding height above the enclosure wall. A further link between the Homeric palace and the Roman temple is the position of the altar: it stood between the gate of the outer wall and the megaron in the palace, and between the gate of the enclosure wall and the sanctuary in the temple.

Although the resemblances between the plan of the palace and the plan of the temple cannot be pushed too far, the Greek temple (and hence the Roman temple, because “the temple is the most Greek of all Roman buildings”) – the Greek temple is certainly the logical development of the simple dwelling house and the hall of the larger house. A very interesting house of the Hellenistic period has been excavated at Priene, in Asia Minor. The high surrounding wall had an opening on one of the short sides, which led to a large courtyard. Across the courtyard, facing the entrance, was the chief room of the house. It was entered through a porch, which had two columns spaced between the side walls. Several Roman temples in Lebanon reproduce these features of the Hellenistic house at Priene; in particular, the temples at Kalat Fakra and Kasr Naous. The ground plans of the temple at Kalat Fakra, and those of the house at Priene and a palace at Troy, are shown in Figure 1. But it is repeated, the resemblances should not be pushed too hard: the influence of Semitic culture and tradition on Roman buildings in the Syrian province must not be overlooked.

With the exception of the small circular temple at Baalbek (Pl. 48-51), the Roman temples in Lebanon are of three types: the antae, the prostylos, and the peripteral. A good example of the antae temple, the Roman templum in antis, is that of Ain Harsha. The side walls extend the full length of the podium, and form the corner supports for the beams and the roof. Two columns, spaced between the side walls, provide the centre support for the beams and roof; these columns lend considerable dignity to the entrance façade. At Ain Harsha the columns have gone, but their bases can be seen clearly (Pl. 1).
In the prostylos temple, the porch is lengthened and the two columns noted in the antae temple are brought forward beyond the line of the side walls. Two more columns provide the corner supports for the beams and the roof, and an architrave joins these columns to the pillars of the side walls. Bziza (Pl. 2) is a typical prostylos temple, but the increased depth of porch is best seen at Kasr Naous, where additional columns in line with the side walls increase the area of the porch to about one third of the area of the entire building (Pl. 109). The prostylos was the Romans’ favourite temple form.

The peripteral temple, perhaps the most perfect of the temple forms, has columns ranged round the four sides of the building to provide a peristyle, or colonnade. The podium is widened to carry the columns which run parallel to the side walls, and the number of columns in the front is increased to six or eight to span the extra width of the podium. The best preserved peripteral temple is at Baalbek, where the north and west sides of the peristyle are still intact (Pl. 3). Apart from the adjoining Temple of Jupiter Heliopolitan, where the famous six columns stand, no other peripteral temple in Lebanon has retained the columns stand, no other peripteral temple in Lebanon has retained the columns of its peristyle.

A word must be said about the orientation of the temples. Almost all the Roman temples in Lebanon are oriented so that the sunlight can enter them. A great many are “oriented” in the true sense of the word: they face due east. A shaft of light from the rising sun can pass through the door to bathe the cult statue in colour. The door, it should be noted, is never blocked by the columns of the porch; sometimes, indeed, the columns are unevenly spaced so that the doorway may be unobstructed. Although most temples in Lebanon face the east, there are some temples which face south or west. Temples facing south are found only in conjunction with larger, more important temples oriented to the east. There is, for instance, a small temple at Niha (Pl. 10) set at right angles to the Temple of Hadaranus; and another at Hosn Niha which lies almost athwart the larger temple (Pl. 15). Temples facing west are rare (Deir el-Kalaa is the most notable), and where they occur it is the setting sun which lights up the cult statue in the sanctuary.

The orientation of the temples in the Mount Hermon region is particularly interesting. It has been asserted that the Roman temples which circled Hermon were oriented to the cone-shaped tip of Kasr es-Sebayb, the highest point of the mountain and the site of a sacred enclosure in Roman and pre-Roman time. Archaeological fervour still blows hot and cold for this opinion. The temples which circle Hermon are, from the south: Hebbariya (Pl. 56), Ain Harsha (Pl. 1, 57 and 58), Ain Libbaya, Nebi Safa (Pl. 59, 60 and 61), Akbeh (Pl. 62), Aiha, Beka (Pl. 63 and 64), Khirbet el-Knese (Pl. 65 to 69), Yanta (Pl. 70 and 71), Deir el-Ashayr (Pl. 72 to 75), Rahle, Burkush, and Er-Rime.

Now the first point to make is that several of these temples do not front Mount Hermon. The four most northerly temples --- Deir el-Ashayr, Beka, Khirbet el-Knese (two temples), and Yanta --- face east or south-east, almost turning their backs on Hermon. Ain Harsha, Akbeh and Nebi Safa face Hermon, it is true, but since they lie to the west of the mountain this was inevitable if they were to be oriented to receive the rays of the rising sun. There is, I think, decisive evidence that it was to the east and not to the summit of Mount Hermon that these three temples were aligned. First, although the temple of Nebi Safa commands a view of the entire flank of Hermon, the doorway of the temple does not face the Kasr es-Sebayb peak. The bearing of the side walls is at least thirty degrees off this line (Pl. 59 and 61). Second, the side walls of Akbeh temple --- also built on a hill from which there is an unbroken vista of Hermon --- point not to the summit but to the northern end of Hermon, to a stretch of the mountain that is masked by an intervening ridge (Pl. 62). Clearly, then, orientation to Hermon was a matter of no great significance to the builders of these two temples, for had they wished an alignment with the summit they could have achieved it without adding a single difficulty to their task. Third, the temple at Ain Harsha faces due east: again an error of about thirty degrees if it is argued that Ain Harsha temple is oriented to the summit of Hermon. The fact is that the main bulk of Hermon (including the Kasr es-Sebayb peak) is not even visible from the doorway of Ain Harsha temple. I have no doubt, then, that temple orientation towards Hermon is fortuitous rather than intentional.

The opinion that it is intentional seems to derive from a suggestion thrown out by Dr. Edward Robinson, the biblical scholar, who traveled in the Hermon region in the summer of 1852. Robinson noted temples at Hebbariya, Aiha, Deir el-Ashayr, Nebi Safa and Rahle. At the last-named site, on the eastern side of Hermon, Robinson said of the ruin that “the front was westwards, towards the snows of Hermon’’, and that “at the eastern end was a semi-circular projection, like that in Greek churches”. From this report, it seems, the theory of orientation towards Hermon evolved. Now, there is indeed an ancient building at Rahle which faces “the snows of Hermon’’, but it is a Christian basilica, not a Roman temple: the small Roman temple, which is remarkable for the apsidal form of its adytum, lies to the north of the basilica. Its ruins were noted but not described by Robinson.

The question of orientation leads to another question: what was the purpose of these buildings? It must be made clear that the Romans did not use their temples as we do our churches: the worshippers did not gather inside a Roman temple. The division of our churches into nave and chancel so closely parallels the division of Roman and Greek temples into cella and adytum that we are apt to think of the temple, like the church, as congregational. This is an error. The temple was built to house the cult image, not to house the worshippers. Priests, augurs, and privileged persons crossed the threshold of the temple; the worshippers did not.

This fact explains the position of the main altar, and explains also the need for an enclosed area when the temple did not front a forum. The altar, dedicated to the deity to whom the temple was consecrated, stood in the courtyard in front of the steps leading up to the porch (Pl. 21 and 36). The officiating priests and the sacrificial assistants stood at the altar; the worshippers filled the space between the altar and the enclosure wall. All faced the doorway of the temple, looking towards the cult image within. In the absence of a paved area (for example, a forum) adjacent to the temple, the enclosed courtyard was needed to accommodate and shelter the worshippers. The position of the altar ensured that these worshippers faced, if not the image of the deity, at least the entrance to the deity’s sanctuary.

I think that in most Roman temples in Lebanon, the climax of the sacrifice on the great feasts must have coincided with the moment when the sun lit up the cult statue. The priest officiating at the altar could look up to the porch and see, through the open doorway, the dim interior of the temple. The rays of the sun would slant through the doorway, and the priest - though perhaps not the assembled worshippers - would see the image of the deity bathed in light. This, I believe, was the moment at which the stroke dispatched the sacrificial victim. The creature for sacrifice would, if small, have been driven to the altar loose; if large, it would have been led to the altar on a long rope: a short halter suggested an unwilling sacrifice.

It is clear, then, that the orientation of the temple is an essential element in the sacrificial ritual. Yet, as was stated above, not all temples face due east. Is it conceivable that the builders of these temples made gross errors in orientation? Surely not. On the contrary, the fact that the axis of a temple may be some degrees off due east suggests exact plotting to me. I believe that it is evidence of a desire to direct the line of the temple to the point on the horizon where the sun rose on the feast day of the deity to whom the temple was dedicated. This, I think, explains the considerable variation in the side bearings of the temples: there are corresponding variations in the position of the sun at sunrise during the course of the year. A temple oriented to sunrise on June 21st will not have the same bearing as one oriented to sunrise on, say, September 1st.

From this conclusion it should follow that the feast of at least one major deity fell on the day when the cult statue in temples with due east orientation received the first rays of the sun. But even with this theory to work upon, few of the temples in Lebanon can be confidently assigned to a particular deity. There is the evidence of the ancient writers for the dedications of some temples, e.g. Venus at Afka; and there is epigraphical evidence for the dedications of others, e.g. Jupiter Heliopolitan at Baalbek and Deir el-Kalaa, Atargatis at Kalat Fakra; but conclusive evidence to identify the principal deity of most of the temples is still lacking. The goddess Nemesis is linked to Makam er-Rab (or Beit Jallouk, as it is known locally) by an inscription and an altar found at the site, yet this evidence is insufficient to identify Makam er-Rab as a temple of Nemesis. Even the much-studied temples of Baalbek defy positive identification: the great temple may with some confidence be assigned to Jupiter Heliopolitan, but the identity of the principal deity worshipped in the small temple and in the round temple remains in doubt.
What is certain, however, is that the god Jupiter Heliopolitan (or his Semitic counterpart, Hadad) was worshipped at many of the Roman temples in Lebanon, and hence his representations at Baalbek and Deir el-Kalaa are of particular interest (Pl. 38). As charioteer of the sun, the god is depicted with a whip in his right hand, and since he is also a storm and rain god he holds a thunderbolt and ears of corn in his left hand. In an exceptionally sunny, well-watered and fertile area such as Lebanon, the insignia of whip, thunderbolt and ears of corn seem particularly apt. No less apt are the bulls which flank Jupiter Heliopolitan, symbolising the god’s fertilizing power. Worshippers at his sanctuaries may well have seen a parallel between the rumbling thunder, which so frequently accompanies the fertilizing rain, and the virile bellowing of live bulls. The god’s cuirass, or body armour, is paneled with the busts of the deities: Sol (Helios), Luna, Mars, Mercury, Jupiter, Venus, Saturn. These are the gods, it should be noted, who presided over the days of the week, and the symbolism of the panels of the panels of the cuirass points to the ancient preoccupation with astral lore.

The Roman temple sites in Lebanon fall into three main groups. First, there is the group in the Bekaa valley north of the Shtaura-Damascus road. Second, there is the group in the area south of the same road, including the Wadi Taym and the western flank of Mount Hermon. Third, the group in the area west of a line drawn along the ridge of Mount Lebanon. These are not geographical areas, of course; there is no quirk of geography to justify putting the temples at Kafr Zabad in the first group and the temple at Majdal Anjar - just a mile or so to the south - in the second. But the road from Shtaura to Damascus does cut the Bekaa valley into convenient halves, and it is useful to have the groups in mind when planning to visit more than a single temple. Hence, with the exception of the first three plates, which illustrate the Roman temple forms described above, the plates in this book are grouped according to the three areas. The index not only lists the plates and the temples cited in the text, but gives details of the route to each temple site. Since Aley – Sofar - Dahr el-Baydar - Shtaura is the quickest approach to the Bekaa valley from Beirut, the routes for all sites in the first and second groups begin at Shtaura.

It will be remarked that the coastal plain of Lebanon is singularly lacking in temple remains, but it must not be thought that the principal coastal cities went unembellished during the Roman era. Berytus, Byblos, Sidon, Tyre, Tripolis, Botrys (Batroon), Caesarea ad Libanum (Arka), were all prosperous enough to have mints and to strike coins under the Romans. There were certainly temples in all these cities; the coin types are sufficient evidence for this. The reverse of a coin of Berytus, for example, illustrates a temple of Astarte (Venus); this coin was struck during the reign of Caracalla (A.D. 211-217), and has the bust of his mother, Julia Domna, on the obverse side. A similar temple appears on the Byblos coinage, and this city struck under Macrinus, the successor to Caracalla, the well-known coin depicting a temple precinct and courtyard built round a baetyl, or sacred cone. But in these urban surroundings, ashlars and column drums were too useful to lie unused; from the Byzantines to the Ottomans, temple debris - particularly the dressed blocks - was utilized in buildings. Even a cursory examination of the medieval fortifications along the coast - at Byblos, for instance - will reveal the extent of the pilfering from Roman buildings. Door frames, lintels, architraves, even altars and inscribed stelae, can be seen in the lower courses of castle and church walls (Pl. 103).

Finally, a comment must be made on the uniformity of form, style and decoration in Lebanese temples. A period of about two hundred and fifty years - roughly from the accession of Augustus to the death of Philip the Arab - spans the foundation and completion of all the temples. Despite the length of the period, Lebanese temples show no fundamental changes in design. The form of the adytum, for instance, is almost unvarying: a flight of steps leads up to a dais where the cult image stood, sheltered by a canopy or framed in a niche. (The small temple of Atargatis at Kalat Fakra (Pl.95) is an exception: in this temple the floors of the adytum and the cella are on the same level. An inscription dates the temple fairly conclusively to the latter part of the first century, but the purpose of the benches in the adytum, and the niches under them, has yet to be explained.) There are no apsidal adytums in Lebanon, as at Rahle, on the Syrian side of Mount Hermon. There are no quadrangular niche adytums, as in the Temple of Bel at Palmyra. With such uniformity, then, it is not difficult to see the hand of a single master builder in the temples; a “Minister of Works’’, as has been suggested in connection with town planning and civic building in the provinces of Rome. Whether this be true or not, the Roman temples in Lebanon constitute a magnificent testimony to the unifying influence of Rome. They provide inexhaustible and absorbing material for the historian, the archaeologist, the architect, the astronomer, and the humble inquiring tourist.

Sketch Map of Roman Temple sites of Lebanon
Based on the map in Romische Tempel in Syrien

Decree N. 2385 of 17/1/1924 as amended by law N. 76 of 3/4/1999 ( articles 2, 5, 15, 49 and 85 ) lays down as follows: The author of a literary or artistic work, by the very fact of authorship, has absolute right of ownership over the work, without obligation of recourse to formal procedures . The author will himself enjoy the benefit of exploitation of his work, and he possesses exclusive rights of publication and of the reproduction under any form whatsoever. Whether the work in question comes under the public domain or not those persons will be liable to imprisonment for a period of one to three years and to fine of between five and fifty million Lebanese pounds, or to either one of these penalties, who 1-will have appended or caused to be appended a usurped name on a literary or artistic work; 2-will have fraudulently imitated the signature or trademark adopted by an author, with a view to deceiving the buyer; 3-will have counterfeited a literary or artistic work; 4-or will have knowingly sold, received, or put on sale or into circulation a work which is counterfeit or signed with a forged signature. The punishment will be increased in the event of repetition.