¹ In a paper entitled ‘The rise of Ottoman historiography’ presented to a conference held at the School of Oriental and African Studies in 1958 and thereafter printed in Lewis, B. and Holt, P. M. (ed.), Historians of the Middle East, London, 1962, at p. 159.Google Scholar

² Namely, the ‘Anonymous Chronicles’ edited by Giese, F. (Die altosmanischen anonymen Chroniken, I, Breslau, 1922Google Scholar, and II (translation) in Abh.für die Kunde des Morgenlandes, XVII, 1, 1925, 1–170Google Scholar) and the chronicle of Uruc (Oruç) b. ‘Ādil (ed. Babinger, F. from the MSS of Oxford and Cambridge, Hanover, 1925Google Scholar; another MS at Manisa, two more at Paris), whom Dr. Beldiceanu-Steinherr has now proved to have been living in Edirne in the middle years of the reign of Bāyezīd II (BSOAS, XXXIII, 2, 1970, 359–63).Google Scholar

³ See, e.g., Lewis, and Holt, (ed.), Historians of the Middle East, 157–9 (İnalcik) and 170–2 (Ménage)Google Scholar. The lists incorporated into these chronicles are similar in style but quite different in content from the ‘royal’ lists, compiled in the Palace, which have survived in independent form (for them, see Ménage, V. L., Neshrī's history of the Ottomans, London, 1964, 15).Google Scholar

⁴ It was Lot 237 in Sotheby's sale-catalogue of that date (Bibliotheca Phillipica, New Series: medieval manuscripts, pt. IV), having been MS 871 in the collection of the eccentric bibliophile Sir Thomas Phillipps (1792–1872). It had earlier been in the possession of Auguste Chardin, being Lot 2628 in the Chardin sale of 9 February 1824 in Paris. It now bears the number 160 in Mr. Digby's private collection. I am most grateful to the present owner for allowing me to describe the contents and to publish a section of the text.

⁵ The text begins, after the besmele, etc.: Ammā, ba'd: bu risālenüñ ṭāliblerine ve nāẓirlarina i'lām olinan budur-ki bir cemā'at iḫvān-i ṣādiḳ ve ḫullān-i yārān-i muvāfiḳ bu az'afi 'l-'ibād-i bīçāreye iḳdām gōsterüb eyitdiler: Her fuṣaḥā ve bulaġādan işbu ‘ālem-i fānīde def'-i melāl ve ḳaṭ'-i sü'āl içün bir yādigār komişlardur; eyle. olsa … and ends: … dōnüb girü İstanbula, geldi. Va'llāhu a'lem. Rāvī eydür: ol zamānda t¯rīḫ seksen altisindayidi. Bundan so¯ra Sulṭān B¯yezīd Ḫān Edrenede bir 'ālī'imāret ve bir medrese ve bir tīmārḫāne ve bir ḥamm¯m ve bir kōprü yapdurdi [ve] müsāfir ve ġurebā ve fuḳar¯ ve ṭalebe rāḥat olmaġiçün. Va'llāhu a'lem. There is an interesting bibliographical peculiarity at fol. lr. This page is filled with text, in the same hand as the rest of the MS, but this text is not (as Sotheby's cataloguer believed) ‘the end of the previous chapter’ of the work beginning on lv; it reproduces the text found on fol. 17r, beginning at exactly the same point. The only explanation is that the copyist was working with unbound 16-page gatherings before him: leaving lr blank, he filled his first gathering (lv–8v), and then his second (9r–16v); but then, instead of taking up an entirely blank gathering, he picked up the first in error and continued copying onto fol. lr, not noticing his mistake until he had filled the page, whereupon he re-copied the continuation from 16v onto the present 17r.

⁶ fī ḳaryat 'l'sbk, with an incomprehensible vocalization; possibly (al-)Esbek or (mis-spelt) İlyās-beg. A copyist of this name (which is, however, of course common) completed a MS of Aḥmedī's Tarwīḥ al-arwāḥ (‘harekeli nesih’) in 932/1526 ‘at Bozdağ’ (Karatay, F. E., Topkapt Sarayi Müzesi kütüphanesi: Türkçe yazmalar kataloğu, 1, Istanbul, 1961, no. 1768).Google Scholar

⁷ Chardin's?

⁸ Fully described by Wittek, P., in MOG, II, 1–2, 1925, 151–64Google Scholar, and evaluated by Giese, F., Die altosmanische Chronik des ‘Āšiḳpašazāde, Leipzig, 1929, Einleitung, pp. 8–12.Google Scholar

⁹ One probative passage is 'Āpz., § 67, the first section of which (ed. Giese, 66.22–67.7) is a patent—and clumsy—interpolation into the text of the Anon. (ed. Giese, 34.23). The Digby MS (fol. 50v) has the same presentation.

¹⁰ The following (fol. 13r = 'Āpz., § 17) is typical: bir cenk-i'aẓīm ėtdiler ki kāfirlere dünyāyitenk ėtdiler, şōyle kim küştelerden püşteler oldi, adam kant seyl oluban yürüdi, ḫūn biḫāri havā yüzin bürüdi; boz at, kara at yek-renk oldi; ak sakal, kara sakal hem-renk oldi; şīrāne ve delīrāne bir cenk ėtdiler kim feleklerde melekler taḥsīn ėtdiler. Āḫir…

¹¹ fol. 57r. Cf. Oruç (ed. Babinger, 37.8–10,105.12–14) and MS A of Neşri (ed. Unat-Köymen, , I, p. 362).Google Scholar

¹² fol. 69v: Rāvī eydür: bu cānibde meger Sulṭān Murāduñ bir oġli var-idi, adina Sulṭān 'Alā-dīn dėrlerdi, Amāsiyye taḫtinda otururdi; ol Sulṭān 'Alā-d¯ne aġu vėrüb helāk ėtdiler. Cf. İnalcik, H., Fatih devri üzerinde tetkikler ve vesikalar, I, Ankara, 1954, 60.Google Scholar

¹³ e.g. cenk for Canik, arġur for Uzġur.

¹⁴ I use the following abbreviations: Giese = Giese, F., Die altosmanischen anonymen Chroniken: Teil 1, Breslau, 1922Google Scholar; (using Giese's sigla) W₁, W₃, M₂ = (respectively) the expanded version of the ‘anonymous chronicle’; the unexpanded but extended version; the Munich MS frequently cited in the apparatus, which is in fact, up to Giese 66.11, a W₂ text and thereafter an ‘Oruç’. E = the Edirne MS of the ‘anonymous’, substantially a W₃ text but with some distinctive (and apparently archaic) features. O, C, M, P = the Oruç texts, (respectively) the MSS of Oxford and Cambridge, published by Babinger, the Manisa MS, and Paris, MS supp. turc 1047. My conclusions concerning the interrelation of these texts are set out in BSOAS, XXX, 2, 1967, 314–22.Google Scholar

¹⁵ The text being unreliable, it is profitless to scrutinize these figures. It is perhaps worth noting that they correspond broadly with those given by ‘Āşiḳpaşazāde (§ 170; Atsiz, , p. 248Google Scholar), who calls his source for them aṣil tevārīḫ (or aṣl-i tevārīḫ?)—possibly a list similar to this. Similar figures appear dispersed through the text of Oruç, and not only in the late recension which avowedly contains interpolations from 'Āpz.: see Buluç, S., Untersuchungen, Breslau, 1938, esp. p. 49.Google Scholar

¹⁶ Read, with Giese, 17.6, 758.

¹⁷ Read, with Oruç, (O 20.20), 760.Google Scholar

¹⁸ MS: kardaşlart ile girü bu yil cenk ėtdi karindaş, with girü bu yil and karindaş deleted (the copyist having slipped into transcribing the next year's entry).

¹⁹ MS: 'rnkmwdn, vocalized? Erenkmüdün, presumably to be emended to İznikde (as EOCM), but cf. P: Er.nkūmūr yöresinde buluşub uġraşdilar.

²⁰ The blinding of ‘İvāz Paşa is mentioned in OP, sub anno 828, in what appears to be an interpolation (see below). That Saruca became vizier as early as 826/1423 is a new detail, certainly textually (if not historically) correct, cf. next annal: hitherto the earliest reference to Saruca as ‘vizier’ has been for 1428 (Ducas, , 205Google Scholar; cf. Wittek, P., Byzantion, xxi, 1951, 328).Google Scholar

²¹ For ‘seized and attended to [i.e. killed]’ Cüneyd, the chronicles have ‘expelled’, recording his death and the taking of Ayasoluk (O, but MP: İpsala) sub anno 829.

²² ECM have substantially the same text (CM adding çevre ėllere akin vėrdi); OP, however, have a long entry recording the campaign against İsfendiyār (spring 1423 = 826) followed by the marriage (different wording), the siege of Constantinople (summer 1422 = 825!), the Küçük Muṣṭafā affair (repetition!) and the blinding of ‘İvāż Paşa (P: İzniki fetḥ ėdüb ol seferden gelüb vezīri Ḥācī ‘İvāż Paşadan bir ḥīle duyub Gelibolida dutub gōzlerini çikardi), the death (P: vefāt ėdüb, O is corrupt) of ‘Laz-oġli’ (Stephen Lazarević, d. 19.7.1427 = 830), and the Hungarian attack on Gügercinlik/Golubac (spring 1428 = 831). It seems that items from another (undated?) taḳvīm have been interpolated here (cf. 0 48.5 and P 38v.9: ba'zilar eydürler/dėrler). To make the textual problem still more complex, the W₃ texts lack the ‘marriage’ item altogether.

²³ MS: Sefer (as CP), but to be corrected to Sakar (with EOM): the Sakar hills lie some 50 km. north-west of Edirne.

²⁴ The chronicles have Laz-ėline and do not name the leaders. The item presumably refers to the operations led from Üsküb (İsḥāḳ Beg's base) against Novobrdo (which was besieged) and Kruševac/Alacaḥiṣār (which was taken), see Jorga, , GOR, 1, 395Google Scholar; Jireček, , Geschichte der Serben, II, 164.Google Scholar

²⁵ EMP and W₁ (but not OC) name also Cān Adasi (W₃: Siçan Adast), presumably the ‘ungarische Insel’ of Jireček, , II, 160Google Scholar. Incidentally, Jireček's suggestion that this was Neu-Orsova = Adakale receives support from the fact that when the Ottomans re-took Adakale (not yet so named) in 1691 ‘it was given the name Şāns Adasi’ (Siliḥdār, , II, 542Google Scholar); no doubt this name refers (as A. Decei points out, in EI, second ed., s.v. Ada Ḳal'e) to the ‘entrenchments’ (German ‘Schanz’, cf. Redhouse, s.v. şānṣ), but the resemblance between Cān and Şāns suggests that Şāns is also a re-casting, as Turkish army slang, of an older Cān; and this in turn appears to reflect the medieval name ‘Saan’ (see Hertz, A. Z., Archivum Ottomanicum, III, 1971, p. 171, n. 4).Google Scholar

²⁶ i.e. ḳāḍī'asker (as P, which alone of the chronicles mentions this item), in succession to Ḫalīl Paşa, who was promoted to the vizierate from that office (OP).

²⁷ Only in P, which also appends the viziers: Ḫalīl, Meḥemmed, Saruca.

²⁸ EOP read Çōkeye (C lacks the ‘yayla’ item), presumably correctly: cf. the letter of Murād II, of 9 June [1431], written na Čoku na planinu (Truhelka, , no. 3, cf. İst. Enst. Dergisi, I, 1955, 43Google Scholar), cited by Babinger, , Aufsätze, 1 (Geburtstag), 169Google Scholar. Babinger could not then (1949) identify the locality; but in his history of Meḥemmed II (Eroberer, 7 = Conquerant, 21 = Conquistatore, 33), without giving a reference, describes it as lying north-west of Edirne.

Administratively, Çöke was a nāḥiye of the ḳażā of Edirne, embracing (with other localities which I cannot trace on the map) the villages of Demirhanh and Ömerobasi (Barkan, Ö. L., in Belgeler, III, 5–6, 1966, 320, 329, 363Google Scholar), the first of which is shown on modern maps some 15 km. east-north-east and the second 40 km. north-east of Edirne. Other villages of the nāḥiye were Karayusuflu and Faki-deresi (Gökbilgin, M. T., Edirne ve Paşa Livâst, İstanbul, 1952, 326, 497Google Scholar), the residents of the last being derbendcis, guarding the road ‘from Bogdan and Dobruca’. ‘Karajusuf’ is marked 18 km. north-east of Edirne on the Austrian General Staff map of 1829. This map also records as ‘Faki-dere’ the river (now Fakiiska, in Bulgaria) on which is situated ‘Umurfakih (Faki)’ (now Fakiya), 70 km. north-east of Edirne on the road to Aitos. The yayla of Çöke is therefore to be located north-east of Edirne, somewhere to the west of the yayla of Keşirlik.

²⁹ MS (clearly): īdī; EOC (sub anno 835): yeñi; but P: iri. The issue of 834 represented a temporary return to the heavier akça of and even ḳīrāṭ (Akdağ, M., Türkiye'nin iktisadî ve içtimaî tarihi, I, Ankara, 1959, 394, 422–5Google Scholar; cf. OCP, with the note that 260 akças were struck from one lidre of silver), so that iri is a tenable reading.

³⁰ E: no name; OCP: Sakar.

³¹ Some words lost by haplography, cf. E: Arnavud vilāyetine akin eyleyüb rāst gelmeyüb Arnavud viläyetinde.…

³² MS: b'rdşky, corrected after E (OP: B.r.zş.k), and cf. Giese, 113.8. For this name see Babinger, F., Aufsätze, I (Elbasan), 202.Google Scholar

³³ The missing words refer to the suppression of the Albanian rebellion (discussed by İnalcik, H., in Fâtih ve İstanbul, I, 2, 1953, 164–6).Google Scholar

³⁴ This entry presumably refers to Sigismund's attack in spring 1428 (Jireček, , II, 164Google Scholar), i.e. 831 (recorded in OP in an interpolation sub anno 828, see above, p. 574, n. 22). ‘Āşiḳpaşazāde, it is true, describes (§ 106) an attack in 837 (and is followed by İ. Uzunçarşili, H., Osmanli tarihi, I, second ed., Ankara, 1961, 414Google Scholar), but his chronology (even his sequence of events) is much distorted for these years; he may well be describing the events of 1428.

³⁵ MS: k.stūrl.k, corrected after EOMP. The locality Keşirlik is marked on modern maps 55 km. north-east of Edirne.

³⁶ There is no record of this raid in the chronicles, under this or another year.

³⁷ This item is not in E or Giese; the Oruç texts give extra detail.

³⁸ MS: būr^uc, apparently understood as Burc in the chronicles. I take the reference to be to Boraç (on the Gruža, south-west of Kragujevac), which was taken, however, according to Jireček, (II, 174–5)Google Scholar in 1438 (= 842–3).

³⁹ Not recorded in E, Giese, or P. OCM (and M₂) mention it earlier in the annal, together with the ‘New’ (i.e. Üç Şerefeli) Mosque.

⁴⁰ MS: todilar tārīḫ anlara ḥummār ad. P has only the second couplet: Ungurusa geçdügi Sul ṭān Murād / komişlar tārīḫin anuñ ḫummār ad. M margin has simply: tārīḫ ḫummār 841.

⁴¹ All texts have the Semendre item, and under 842: the campaign, beginning in March 1439 with Semendre falling in August, covered the two years 842–3. OCP record ‘All Beg's expedition (not mentioned in E or Giese), but in different words, sub anno 843, immediately after the Belgrade item—and perhaps correctly, if it was intended as a diversion during the siege of Belgrade (cf. Jireček, , II, 176Google Scholar), i.e. in summer 1440.

⁴² The ‘ḥavāle’ point appears in OP (only) after the mention of Şehābeddīn Paşa sub anno 846.

⁴³ ECM give no names, O has two, and P (adding Ḫiżr Bah) has four.

⁴⁴ MS: dėmişler anuñ tārīḫin zevāḳ-i mā, emended after P (47v.16), which alone of the chronicles gives this couplet.

⁴⁵ Sic; chronicles: Maġnīsāya. A few words have perhaps been lost.

⁴⁶ This obit appears only in P.

⁴⁷ P gives (52r.4), with no couplet, the chronogram ḫāliḳ-i subḥān (either phrase gives the correct total, 852).

⁴⁸ At this point the chronicles give the date 852 and begin a new annal; see below, p. 580.

⁴⁹ Not ‘the menāḳib of Yaḫşi Faḳīh’, see BS0AS, xxvi, 1, 1963, 52.Google Scholar

⁵⁰ When I give no specific reference for a date I am following the chronology in Halil İnalcik's invaluable critical collation of the Ottoman and the Western sources in his article ‘Murad II’ in İslâm ansiklopedisi.

⁵¹ H. İnalcik, in Lewis, and Holt, (ed.), Historians of the Middle East, 159.Google Scholar

⁵² The exception is the system of dating, for the discussion of which see below, p. 583.

⁵³ E 34v.11 = Giese 67.3; 0 53.24 = C 117.18 = M 45v.15 = P 43v.8.

⁵⁴ -iken / bu e nāda: E 38v.4 = Giese 70.26; O 59.24 = C 121.20 = P 49r.3.

⁵⁵ Some of these items are considered below, p. 582.

⁵⁶ E 31v.3–32v.3 = Giese 56.2–59.16; O 46.6–47.15 = C 112.11–113.15 = M 41v.4–42v.12 = P 37r.4–38r.6.

⁵⁷ §§ 83–7.

⁵⁸ It is interesting that the Annalist regards Muṣṭafā as Murād's ‘brother’ (in fact he was his uncle) and yet calls him ‘Düzme’, ‘Impostor’. The ‘brothers’ of the annal for 824 must be the two Muṣṭafās, for the haggling with the Byzantine Emperor over the two younger brothers, Yūsuf and Maḥmūd, hardly ranks as ‘warfare’ (cenk). This tradition of the ‘two brothers’ seems to have figured also in the chronicle-source: see ‘Āpz., heading to § 82 (ve karindaşlari daḫi ne oldi, Yūsuf and Maḥmūd not being mentioned until § 94), and cf. Giese 55.29: Sulṭān Meḥemmedüñ oġlanlari her ṭarafdan baş kaldurdilar (so too E, whereas the phrase is lacking—edited out ?—in Oruç). ‘Rūḥī’, incidentally, also refers to Muräd as Düzme Muṣṭafā's ‘brother’ (MS Marsh 313, fol. 106v, but omitted in the Algiers and Berlin MSS).

⁵⁹ O 64.4—16 (and now P 52r.5–19). C does not mention the marriage.

⁶⁰ Aufsätze, I (Heirat), 229–32.Google Scholar

⁶¹ Fatih devri üzerinde tetkikler ve vesikalar, I, Ankara, 1954, 108–9.Google Scholar

⁶² cf. also ‘Āpz.'s moralizing comments in § 121, that with his son married Murād had achieved all his earthly goals.

⁶³ For the historical context (the installation of a pro-Ottoman voyvode in Wallachia) see İnalcik, H., Fatih devri, p. 98, n. 130.Google Scholar

⁶⁴ The redactor of W₁ (who revised and expanded the text, not always successfully) gives the apparently impeccable sequence:

853: Giurgiu; raid into Wallachia and appointment of new voyvode;

854: siege of Akçeḥiṣār; marriage ‘that winter’;

855: death of Murād, Wednesday 1 Muḥarrem.

It seems that he had access to a good source (Enveri?), which gave him the ‘Wallachia’ item and the correct sequence.

⁶⁵ The strong evidence in favour of 835 is summed up by İnalcik, in Fatih devri, p. 55, n. 2.Google Scholar

⁶⁶ Enverī, the only other Ottoman source to give a date, has 830; cf. Fatih devri, p. 55 and n.4.Google Scholar

⁶⁷ cf. also the tone of ‘Āpz.'s § 105.

⁶⁸ See above, p. 576, n. 34.

⁶⁹ ‘Āpz. accompanied İsḥāḳ Beg on this pilgrimage (§ 113), but his chronology is in confusion. However, since he states that they returned ‘before Semendre fell’ (August 1439), the pilgrimage they made was presumably that of 841 (Dhū 'l-Ḥijja 841 = June 1438).

⁷⁰ He is called ‘merḥūm’ in a vaḳfiyye of 848 (İnalcik, , Fatih devri, p. 83, n. 67).Google Scholar

⁷¹ See above, p. 574, n. 21.

⁷² So Jireček, , II, 178Google Scholar. İnalcik (‘Murad II’, col. 607a), following the chronicles, prefers 1440.

⁷³ A discrepancy of one year in the middle years of Murād's reign is not surprising, for the 31 Hijrī years cover only 30 solar—and thus ‘campaigning’—years and in the decade 835–45 Muḥarrem moved back from September to May; hence even a careful annalist could easily misdate by one a summer's event falling this side or the other of the Muslim new year. A clear example is the eclipse and the comet both recorded sub anno 836: the eclipse occurred on 17.6.1433 (Oppolzer, no. 6284, see Babinger, , in MOG, II, 3–4, 1926, 313Google Scholar), which indeed falls in 836, but the comet (September 1433, see Grumel, V., Traité d'études byzantines. 1. La chronologie, Paris, 1958, 475Google Scholar), though it appeared at the end of the same summer, in fact belongs to 837.

⁷⁴ M 44r.1 (vezīrler meẕkūrlar, apparently referring to 837); 0 51.19 (vezīrleri kemā-kān, immediately after the date 841).

⁷⁵ The naming of the three viziers at Giese 55.24–5 (sub anno 824) is part of the ‘narrative’ and stood in the chronicle-source (cf. ‘Āpz., § 81). (The Annals text, naming only two for 824, is not necessarily defective, since Bāyezīd Paşa had probably been killed by the end of the year.)

⁷⁶ Hence İnalcik, following O, has dated Fażlullāh's appointment to 840 (‘Murad II’, 605bGoogle Scholar); but the Annals' date and wording (ḳaṣd ėdüb) fortify his point (Fatih devri, 1Google Scholar) of the connexion between Fażlullāh's entry into the Dīvān and the adoption of an aggressive policy in Europe.

⁷⁷ See BSOAS, xxx, 2, 1967, 321–2.Google Scholar

⁷⁸ These, and some other early items, are evidently interpolations in the basic chronicle-source.

⁷⁹ Respectively ‘844–5’ for 842–3; ‘838’, ‘846–8’, ‘851’ for 837, 845–7, 850.