¹ Notably, the composition of the labour force, its strength, provenance and relation to the sequence of construction. Cf., Barkan, Ö. L., “L'organisation du travail dans le chantier d'une grande mosquée à Istanbul au XVI siècle,” Annales, Economies. Civilisations, 17 (1962), 1093–1096.Google Scholar

² The idea that the documents required critical consideration arose from correspondence with Dr. Colin Imber of the University of Manchester. I have subsequently profited considerably from his observations and, in particular, from the generous comments and suggestions of Professor V. L. Ménage of the University of London. To both I express my gratitude. I am also indebted for help on specific points to Mr. Godfrey Goodwin, Professor Wheeler Thackston of Harvard University, Dr. Veronica Tatton-Brown of the British Museum, Professor Magdi Wahba of the Arab Academy. and Mr. Raymond White of the National Gallery in London.

³ Goodwin, G., A History of Ottoman Architecture (London, 1971), pp. 215–239, 479–482; nn. 72–165 and references.Google Scholar

⁴ Only partially published. Cf., L. A. Mayer, The Buildings of Qāyt Bāy as Described in his Endowment Deed, Fasc. I (London, 1938).Google Scholar

⁵ See author's note above.

⁶ For example, ibsid/ispit (“felloe”), tarbizon/tirabzan (from Persian dārbazīn/“banisters”). Others are much more intractable for example, balgir/girbal/galbir and iflamur, both used, apparently, for ropes. The abundance of variant readings, mainly, doubtless, scribal errors, makes it futile to establish the primary form. The source, in fact, appear to be (Kahane, H., Kahane, R. and Tietze, A., The Lingua Franca of the Levant [Urbana, III., 1958] s.v.), respectively, *cordellina/(Venetian) “fine cord” and *ligadura/(Venetian) “lashing”.Google Scholar

⁷ Gökyay, O. S., “Risale-i Mimariyye—Mimar Mehmet Ağa—Eserleri,” in Ismail Hakki Uzunşarşili'ya Armağan (Ankara, 1976), pp. 113–216.Google Scholar

⁸ The transcriptions mostly follow the text established by Barkan (Süleymaniye Camii ve lnşaati, Vols. I–II), alternative readings being given when they appear to be an improvement. Three terms are consistently misspelled in the documents: ⊂ālam for ⊂alem/(here) “finial”; a⊂lā for ālā/“brightly coloured”, “variegated”; and ṭilā⊂ for ṭilā⊃/gilding. It is more economical to note the fact here and tacitly correct the occurrences whenever necessary.

⁹ Cf., V. Meinecke-Berg, “Marmorfliesen—Zum Verhältnis von Fliesendekoration und Architektur in der osmanischen Baukunst,” Kunsi des Orients, 8, 1–2 (1972), 35–59. She makes the point that the popularity of such marbled tiles (imitating both porphyry and split-marble panelling) is characteristic of the taste of the later sixteenth century. Marbled tiling in buildings which underwent restoration works or which were constructed in the seventeenth century–the Hirka-i Sa⊂det apartments (built shortly after 1517 but tiled much later, including a restoration of 1017/1608–1609), the Bağdad and Revan köşks in the Topkapi Saray (built, respectively, in 1639 and 1635) and the Sünnet Odasi (restored 1641)—must, therefore, be reuses of older tiles. As for those ordered for the mosque of Sultan Ahmed, she argues, they were for the royal box.Google Scholar

¹⁰ Anhegger, R., “Quellen zur osmanischen Keramik” in Otto-Dorn, K., Das Islamische lznik, Istanbuler Forschungen 13 (Berlin 1941), pp. 165–167, nos. 1–4, publishes excerpts from a series of orders regarding the tile-work of the mosque of Selimiye at Edirne. The first (dated 977/1569–1570) orders tiles from Iznik corresponding to the patterns which had been sent there. No. 2 (dated 8 Muḥarram 980/21 May 1572) names the calligrapher, Mollā Ḥasan, as responsible for the inscriptions in the mosque and orders the mi⊂mār bāşi, Sinān, to indicate to him which parts of the mosque are to bear tile inscriptions and which are to be painted: the order is then to be executed by the kāşī emīni. This seems to leave Sinān ultimate discretion as to how much of the mosque interior should be decorated, as no. 3 (dated 4 Rabī⊂ 1980/15 July 1572) confirms. The mosque is to be tiled up to the [height of the ground-floor] windows [on the kible side], with the tile panels above the windows bearing the Fātiḥa. (Actually, the tile revetment reaches up to the full height of the windows and the inscription panels combined.) The short interval between nos. 2 and 3 may suggest, misleadingly, that the inscriptions were no sooner drawn Out by Mollā Ḥasan than they were executed at lznik. Tile panels bearing the Faāiḥa, however, must have been standard Iznik products, and no. 2 must relate to the monumental inscription frieze running below the cornice on the ḳible wall to either side of the miḥrāb. This must have been a much slower business, and it is probable that it is to these designs/nümüne that the complaint (no. 4, dated 29 Cumādā II 982/16 September 1574) of delay in the execution of the tile-work at Selimiye refers.Google Scholar

¹¹ Though very little of the rich wall-painting of the period has survived in bulk a fragment of console-decoration in the mosque of Sokollu Meḥmed Paşa in gold, shades of blue, pale mauve, white, red and black, complementing the tile-work of the ḳible wall, gives a good idea of the effect which could be achieved (cf., K. Otto-Dorn, “Osmanische ornamentale Wandmalerei,” Kunst des Orients, 1 [1950], 45–54). It is highly probable that such decoration followed closely the evolution of tile-designs. Professor Otto-Dorn has noted that the pattern of the painting fragment in Sokollu Meḥmed Paşa recurs in the tile-work of the small mosque of Tāḳiyeci lbrāhīm (1000/1591–1592), outside the land walls of Istanbul. Further evidence for painting taking on a life of its own is the recurrence of parallel motifs in painted decoration in the Ramazan Efendi mosque (995/1586–1587). But the colour-schemes which survive are notably more restricted than the list of colouring materials given in the defters.Google Scholar

¹² The documents so far published on the naḳḳāşḫāne have dealt mainly with the personnel (cf., Rifki Melûl Meriç, Türk nakiş san⊂ati araştirmalari. I. Vesîkalar, Ankara üniversitesi Ilâhiyât Fakültesi, Türk ve Islâm San⊂atlari Enstitütüsü Yayinlari [Ankara, 1953]).Google Scholar Under Süleyman, Selīm II and Murād III, as in much later periods, they were evidently paid by the central authority (Ḥazinedār), from the naḳḳāşbāşi down to the journeymen-pupils, though not all the documents list the complete staff. Under Murād III in particular there are also records of gifts of Ḥāsene sikkes or il⊂as (presumably, special rewards, not salaries), sometimes with a citation naming a specific manuscript, to the naḳḳāşbāşi or other painters (Meriç, , Türk nakiş san ⊂ati, p. 56, no. CVI). Thus, for the Kitāb-i Slyer-i NebevīGoogle Scholar (Ibid., p. 58, item 20) executed for Murād III towards the end of his reign, the account dated 10 Ḏu⊃l-Ḥicce 1003/21 August 1595 states that the azinedārbāşi received 200 florins, the Kethuda 100, the Naḳḳāşbāşi 40, the Hazine Kātibi (again, a Treasury official) 30, and the head of the painters and illuminators 30. The principles of distribution will be familiar to anyone cognizant with modern honours systems.

But there is another set of documents, the most illuminating so far published (Ibid., pp. 66–68, no. CXV, dated 992/1584–1585) relating to the illumination of a Koran written by, or attributed to (evidently, therefore, after some considerable lapse of time), the famous calligrapher Aḥmed Ḳarāḥiṣārī. Paper (in the circumstances, for mounts) was issued as required at the charge of the ẕazinedār or other high military officials. There follow issues of inks, sizes and pigments, over a series of months, which suggests that the accounts for the execution of the manuscript show day-today needs—doubtless with some time-lag, since the craftsmen's requirements would take time to process. The bureaucratic detail would do credit to the British Civil Service, and the impossibility of strikes must have commended the procedure even more highly to the officials responsible. However, the system must have given the naḳḳāşbāşis, who would have worked far better on a block grant for materials direct to the naḳḳāşḫāne, many quite unnecessary headaches. In fact, even if the recorded sequence of requisitions, against the probabilities, faithfully conveys the illuminators' progress, the operation must have been impossible on such an occasional basis. The naḳḳāşḫāşis, on the contrary, must have drawn up a plan and worked as far as possible to it; the craftsmen worked as they were able; and the administrators may have done what they could to avoid undue interruptions. The records, therefore, presuppose a degree of foresight, which may have run to hoarding materials, excess orders or unnecessary requisitions; but some, if not most, of the memoranda must have been provoked by complaints of delay or bureaucratic hindrance. The entries relating to pigments and other materials most suitable for book-production in the Süleymaniye defters could well, therefore, relate to impromptu attempts to deal with this. Like the lists of salaries to library officials, these documents demonstrate that the working of the library was totally, and dangerously, dependent at virtually every stage upon the central authorities.

¹³ Cited after von Hammer-Purgstall, J., Geschichre des Osmanischen Reiches (Pest 1828), III, 518–522.Google Scholar The diary of Archbishop Verantius/Verancsics, covering his mission to Constantinople in 1567, has been published by Kovachich, M. G., Scriptores minores rerum Hungaricum, Vols. I-II (Buda, 1798), but is too brief to justify Hammer's reference to it.Google Scholar

¹⁴ The vaḳfiye of Süleymaniye (ed. Kemâl, Edib Kürkçüoğlu, Süleymaniye Vakfiyesi, [Ankara, 1962], pp. 109–110, text; p. 36 trans.) appoints sweepers or cleaners for the carpets (bisāṭ, enmāṭ [the latter, evidently, from nemed/felt cloth] and seccāde) of the mosque. The forms are evidently collective: this is, therefore, evidence for the inclusion of floor-coverings in the furnishings endowed upon the mosque. It is no evidence, however, for their luxurious quality. In any case, the use of the term seccade/prayer carpet(s) suggests that, on the analogy of the floor-coverings of the Sultan's box/hünkậr maḥfili in the mosque of Ahmed I (see above page 305), they were small, not large.Google Scholar

¹⁵ “⊂Abd,” “Ghulām,” Encyclopaedia of Islam 2.

¹⁶ Watchmaking is a case in point. The Sultans in the sixteenth and seventeenth centuries, as Kurz, Otto (European Clocks and Watches in the Near East [London], 1975)Google Scholar has brilliantly shown, had a mania for European clocks of indescribable complication. This was supplied by lavish diplomatic gifts, tribute (cf. the Habsburg Türkenverehrung following the armistice of 1547) and imports. In the later sixteenth century, however, Murād III had as his müneccimbai/Court astronomer a Syrian clock-maker, Ṭaḳī al-Dīn (1525–1585), who composed a treatise in Arabic on the various clocks and watches he had made and who built an astronomical clock for the short-lived observatory founded at Tophane in 1579 (cf., S. Tekeli, 16⊃inct asirda Osmanli saat ve Takiüddin⊃in “Mekanik saat konstrüksüyona dair en parlak yildizlari” adh eseri Ankara, 1966]). There was a rapid expansion of watchmaking in seventeenth century Istanbul, but primarily at the hands of European craftsmen, for the taste of the Court remained highly responsive to European fashion and the evolution of new types of watch. Clearly, this demand created a demand among the general public for time-keeping devices, with a local school of craftsmen to cater to it: ultimately they too came to cater for the Court. But it remains, doubtless, true that the extent in general to which the Court workshops influenced public taste in Ottoman Istanbul is a matter of nice judgement, and much further work.Google Scholar

¹⁷ Cf. the frequent complaints in the documents published by Anhegger (Ibid., n. II above) that the craftsmen, instead of busying themselves with the execution of Court orders, were occupied in making pottery for private clients.

¹⁸ The ultimate explanation of the plans for the tile decoration of Süleymaniye and subsequent mosques was, as Arthur Lane persuasively argued, Süleyman's restorations of the Dome of the Rock in Jerusalem in the 1540s (cf., A. Lane, “The Ottoman Pottery of Isnik,” Ars Orientalis, 2 [1957], 247–281),Google Scholar which involved, inter alia, the revetment of most of the exterior drum with tilework. This has been disputed by Denny, W. (The Ceramics of the Mosque of Rüstem Pasha and the Environment of Change [New York and London, 1977], pp. 99ff.)Google Scholar on the grounds that the extant tiles datable to the 1540s are not Iznik work and that the identifiable Iznik tiles are later. However, this does not preclude the probability that the idea of large-scale tilework revetments came to Süleyman from his works in Jerusalem and that the restorations there were initially the cause of the massive expansion of production at lznik. Süleyman's restorations in Jerusalem from the 1530s onwards indisputably affected the forms and decoration of architecture patronised by him in Istanbul (cf., J. M. Rogers, “Innovation and Continuity, in Islamic Urbanism,” in Symposium on the Arab City: Its Character and Islamic Cultural Heritage. Medina, 24–29 Rabī⊂ II 1401/28 February–5 March 1981, I. Serageldin, , ed. [Washington, D.C., in press]).Google Scholar Specific evidence, therefore, must be adduced to show that Süleymaniye was outside this general motivation. Denny's thesis, that the appearance of mass production of tiles at Iznik is a phenomenon of the later 1550s, if not the 1560s, postdating Süleymaniye by a few years, is not borne out by the Süleymaniye defters. They cover too long a period, anyway, to make it possible even to conjecture the precise chronology of the orders; but it is not conceivable that the entries and payments for lznik tiles, whether estimates, impromptu additions or orders towards the end of building operations, would relate to material which was only produced some years after the inauguration of the mosque.