The inscription around the rim reads: bismillah al-rahman al-rahim, ashhad 'an la ilah illa Allah, wa 'an Muhammad rasul Allah (In the name of God, the Merciful, the Compassionate; I attest that there is no god but God and that Muhammad is the Prophet of God).
The door knocker is heavily cast in one piece from a copper alloy. The alloy is mostly copper, combined with zinc, a little tin, and small amounts of silver and lead. The very low level of lead used in the alloy is unusual for most bronze vessels produced in the Islamic world, both East and West. It is however very similar to the alloy of a small number of very large scale figural bronzes, notably the Griffin of Pisa, the lion sold in these Rooms 19 October 1993, lot 293, and the hind also sold in these Rooms 25 April 1997, lot 259. All these bronzes have a minimal lead content, probably only from impurities. In all of them the silver content is almost the same or actually higher than the lead content. Of the four bronzes, the analysis of the content of the present knocker is closest to that of the Pisa Griffin, both having a slightly lower silver content than the other two pieces.
The material used to fill in the background of the inscription is also not the bituminous compound found normally in Islamic bronzes. It is a binary silver-lead alloy which has been reacted with sulphide to form niello. The silvery circles within this black niello are of unreacted alloy caused by the imperfect rection with the suphide. This niello is in contrast to the bituminous compound which was used as the background to the inscription on the large lampstand sold in these Rooms, 20 April 1999, lot 512, which is the only other large bronze of this group to have had a second material included.
As well as the metallurgical composition and scale of conception, this bronze shares many stylistic features with the group of bronzes mentioned above. The similarity of the face to that of the large lion is immediately apparent, particularly in the treatment of the nose and mouth juncture, remembering that the lion was almost certainly inset with semi precious stones which would have rounded off the upper lip more than it appears now. Both pieces also make use of rows of small triangular-shaped engraving following the contours of the face, although in the present piece they do not face the same direction. The scrolls used to depict the mane on the two lions also differ in form, but the small drilled points which are a feature of the mane of the lion and the wings of the griffin, can also be found on all three peices, together with the lampstand, used in some of the lettering, particularly at the base of the line dividing two vertical strokes as in the name Allah. Another feature which links the lettering of this piece to that of the others is the addition at times of upper horizontal extensions of the normal vertical hastae, a feature found on the Pisa griffin and also on the lampstand.
There are also many idiosyncratic design features in the present door-knocker, apart from the use of niello. Notable among these is the slightly crudely cast cockerel's head surmounting the knocker. It is slighltly reminiscent of, but far more ferocious than, the bird which topped the lampstand. The scrollwork which forms the background to the lettering is also very unusual. Islamic scrolls almost always terminate with a split palmette; whether split in two, three or more is the usual question. Here the terminals are rounded as axe blades with only one of their subsidiary terminals being split. The surface of each of these is pounced, presumably to avoid them distracting from the plain metal surface of the inscription. The vine scroll itself however has more structure than those behind the inscriptions of any of the other pieces, meandering steadily from one side to the other as it sends off its tendrils. Then there are the spiral eyes, more reminiscent of Ka in the Disney's The Jungle Book than any other Islamic sculpture, which must imply a mystical dimension to this impressive head.
Door knockers with the ring held in the mouth of a lion are a well known Roman motif, being found in numerous examples on all scales of sizes. In the 10th and 11th centuries within Christian Europe these were again popular. The famous bronze doors of Saint Michael's Hildesheim, dating from around 1015, have as their focal point lion door-knockers after the antique (Lasko, Peter: Ars Sacra, 800-1200, London, 1972, pl.111). The bronze doors of the cathedrals and churches of such far-distanced places as Verona (circa 1135-40), Novgorod (1152), Gniezno in Poland (circa 1175), and Benevento (circa 1200) show the univerality of the motif in the Christian world. The mausoleum of Bohemund (died 1111) in Canosa, Bari, South Italy, has a very interesting added feature. Around the lion's head, which itself is closer to the antique than those mentioned above, is a band of decoration which must derive from a kufic inscription (Gabrieli, Francesco and Scerrato, Umberto: Gli Arabi in Italia, Milan, 1979, pls.337 and 338). We have the same elements as those found here, but with a very different aesthetic. For a brief discussion of the Islamic influences in Southern Italy see Garton, Tessa: "Islamic elements in early Romanesque sculpturen in Apulia", Art and Archaeology Research Papers (aarp) no.4, December 1973, pp.100-116).
Indications of Eastern sources:
Recent excavations in Pakistan lend an interesting new light on the appearance of this form in the arab Islamic world. At the site of Al-Mansurah, about 75km. from present-day Hyderabad in the Sind province, four remarkable door-knockers were discovered on the site of the major mosque (Khan, Ahmad Nabi: Al-Mansurah, A Forgotten Arab Metropolis in Pakistan, Karachi, 1990; Zebrowski, Mark: Gold, Silver and Bronze from Mughal India, London, 1997, pls.3 and 4). These are even more massive than the present example, measuring just over 55cm. in diameter. The central head of each is surrounded by a band of deeply engraved kufic religious inscription, beginning with the bismillah and in at least one of the four cases continuing to the shahada. The execution of these four door-knockers is unlike that of the present door-knocker. The faces are highly stylised, almost human in caricature. Technically also these Eastern door-knockers are very different, the bronze having a high lead content averaging between 16.7 and 18
After the province was conquered by the Ummayyad adventurer Muhammad b. Qasim, his son 'Amr founded the city in 738 AD. The city is recorded as thriving under the Habbari dynasty in the 9th and 10th centuries. By the 11th century it is only mentioned in passing and in the 14th it is recorded as being completely ruined (Friedmann, Y.: "Al-Mansura", Encyclopaedia of Islam, new edition, Leiden, 1991, vol VI., pp.439-440). While their geographical position gave the rulers considerable autonomy, they bore alliegiance to the Abbasid caliphs and, in the second half of the 10th century, also recognised the authority of the more local Buwayhid dynasty. The door-knockers are thought to have been made for the palace of the Habbarid emir 'Abdallah ibn 'Umar (r.883-913). The door knockers produced under his aegis clearly derive from an earlier prototype. It is tempting to think that both those and ours derive from a common presumably Umayyad arab original.
One other feature which has some relevance is the form of the calligraphy. The letters he as in rahman rise up and curve back over themselves. Another oddity which one can only just deduce is that the other he of ashhad brings the central stroke up into a vertical. Both of these are typical features of North African inscriptions and are rarely encountered elsewhere.
Dating and origin
We have seen how the door-knocker relates both stylistically and technically to the group of large bronzes mentioned above. Apart from the claim of Fatimid Egypt, which has now been discounted, the two main areas which have been suggested as the places of manufacture for one or more of them is either Spain or Southern Italy. The South Italian suggestion is based on a study of technical similarities with bronzes which were made there, coupled with the known very active centres of production in other materials, notably in Islamic Sicily. Furthermore, there are three large candelabra, similar in size and some decorative detail to that sold here last year, which can with some confidence be linked to the christian community in Amalfi in the 11th century (Bouras, Laskarina: "Three Byzantine Bronze Candelabra from the Grand Lavra Monastery and Saint Catherine's Monastery on Mount Sinai", Deltion, Tis Chrystianikis Archaiologikis Etaireias, Athens, 1991, pp.19-26). Of added relevance with regard to the present door-knocker, is the use of niello inlay on bronze in the two Amalfitan candelabra on the Grand Lavra on Mount Athos. This is a technique which the author notes is otherwise unknown in the middle Byzantine period.
The Spanish attribution is made principally on the features shared between the group of pieces noted above and items of known Spanish archaeological origin, notably the Cordoba stag, and the Monzon lion in the Louvre. The similarity of the metal composition of the Christie's hind, which due to its similarities with the Cordoba stag must be of Spanish origin, and the other pieces in the group, also reinforces the Spanish attribution, although the hind is of course two hundred years earlier than the others. With these more proven arguments in favour of a Spanish origin, this seems certainly the more probable attribution.