Corrections
A public record of everything we have changed our mind about, and what persuaded us.
When something we have already published turns out to be wrong, or the scholarship moves on, we say so here rather than quietly editing the page. Each entry below names the page and the section it affected, and that section carries a dated link back to this record.
| Corrected | Page | What we said | What we say now | On what evidence |
|---|---|---|---|---|
| Great Theatre | “At the western end, the Middle Harbour Gate opened onto the quay through three passages.” | “At the western end, three monumental gates opened from the quay into the city, and the Arkadiane began at the middle one.” The count is the excavators’ own and is not quite closed: Groh’s survey describes a fourth gate building closing the same street at its eastern end, and an older three-passage gate some 250 m further east, which may mark the Augustan harbour basin before the water moved west. The three gates on the quay are Trajanic to Hadrianic; the middle one was built at the end of the first century or the start of the second. | Forschungen in Ephesos III, pp. 8–9, records the discovery sequence — south gate 1896, middle 1899, north 1902 — and Wilberg’s lowering of the middle gate’s date to the Hadrianic period, against Benndorf and Niemann’s Hellenistic harbour gate. The fourth gate and the gate 250 m further east rest on Groh’s survey alone and are not corroborated elsewhere in print, so they are given as his reading rather than as the count. | |
| Great Theatre | The page already treated the stage building as an explained dispute: AD 85 as the headline date, with the older sequence under Claudius, Nero and Trajan named and given as superseded. | The prose is unchanged, and rightly so. What has changed is the standing of the date behind it. AD 85 is no longer a contested reading but a confirmed one, so the page’s account of the Flavian remodelling as the theatre’s most far-reaching now rests on settled ground. | The building inscription on the architrave of the stage building’s lower storey gives AD 85. Krinzinger and Ruggendorfer 2017, p. 517, and Bier 2011, p. 84, confirm it independently, which makes three sources against Elliger’s AD 66. The older chronology, spreading the stage building across Claudius, Nero and Trajan, is not a rival reading of the inscription; it predates the publication that read it. | |
| House of the Virgin Mary | “A letter from the Council of Ephesus in 431 is sometimes quoted as proof. It mentions Ephesus together with John the Theologian and the Theotokos, but the verb is missing, so it cannot show that Mary lived or died there; the 1909 Catholic Encyclopedia read it as meaning only that Ephesus had a church of the Virgin.” | That sentence stands, and the wider record is added behind it. All eight surviving fourth-century versions of the Transitus Mariae agree that John cared for Mary and that she died and was buried near Jerusalem; none brings her to Ephesus. Cyril of Alexandria’s own letter about the council — one of only two ancient texts that connect Mary to Ephesus at all — describes the meeting place only as a great church that some called Mary’s. | Limberis, in Koester 1995, p. 325, for the eight versions of the Transitus Mariae, and p. 327 for Cyril’s letter to the Egyptian clergy (Letter 24; Mansi 4.1241). Epiphanius, in the Panarion, had already observed that although John was preparing to travel to Asia, scripture nowhere says he took Mary with him. Cyril and Epiphanius are known here through Limberis rather than at first hand. | |
| Basilica of St John | “Its walls have 15 towers and two gates, east and west, and enclose the remains of a palace, a mosque, a bath and cisterns. It is a separate circuit from the lower wall around the church precinct, and the two are easily confused.” | The Ministry of Culture’s count is scoped explicitly to the summit castle, and the second, larger count is then explained rather than left as a contradiction: twenty towers in all, six of them guarding gates, on the walled circuit around the church below — including the four square towers flanking the Gate of Persecution, which belongs to that lower circuit. The two counts most likely describe different walls. | Muss and Bammer 2001, p. 28, describes two Byzantine fortification systems on the hill — an inner terrace wall around the church’s atrium and apse, and an outer wall 2.70–2.80 m thick faced with marble carried up from Ephesus — and counts twenty towers, six at gates, round, pentagonal or square. Foss 1979, p. 113, describes the same spolia-faced wall, over four metres thick, and the Gate of Persecution’s towers rebuilt as pentagonal. Whether that outer wall and the Ministry’s castle are one circuit under two names is unresolved. | |
| Church of Ephesus | Egeria’s date was carried only implicitly, behind the general sentence that Ephesus drew pilgrims above all to John’s tomb on Ayasuluk hill; the 380s, Willibald and Constantine VI were not spelled out in the visible prose. | “Egeria is usually said to have planned her visit in the 380s, the standard dating of her surviving travel diary, though Foss, and later Ladstätter and Magdalino, place her nearer AD 400 — and of the two, only that later pair actually claim she reached the city, rather than merely set out to.” | For the 380s: Karydis 2016, p. 6, dating the pilgrimage 381–384, and Pülz 2010, p. 120. For about 400: Foss 1979, p. 43, and Ladstätter and Magdalino 2019, p. 257. Two voices each way, so the page attributes rather than asserts, and keeps the 380s as the standard dating of the Itinerarium. Willibald came in 724–726 and the emperor Constantine VI in the 790s. | |
| Church of Ephesus | “It is usually dated to the early 5th century, though excavation suggests its dedication to Mary may be no earlier than about AD 500.” | “Exactly when it became a church is one of the most argued-over questions in Ephesian archaeology: it is loosely dated to the early 5th century, excavation puts the standing walls no earlier than about AD 500, and the most radical reading holds that nothing was built there at all until after the 431 council had already met in the bare hall.” The full three-position debate is on the Church of Mary page. | Karwiese, in Koester 1995, pp. 312 and 316, dates the foundation several decades after 431 and the side walls no earlier than about 500; Degasperi 2013, p. 19, follows him; Russo puts building work on the site as early as the first half of the fourth century. Pülz 2010, p. 134, goes furthest — nothing built until after 431, the bishops meeting in the adapted stoa — and supplies the objection to his own case, since the widening of the nave and the new apse would have had to be done between November 430 and June 431. | |
| Around Ephesus | The excavation was described only as a late-antique cemetery with a cave church, several hundred graves of the fifth and sixth centuries and a great many lamps, which left a reader to assume the complex belonged to Theodosius II, the legend’s own emperor. | The building itself is dated archaeologically to the late fourth century, under Theodosius I — half a century before the miracle it is supposed to mark — with the dedication to the Seven Sleepers inserted only in the fifth. The legend keeps its own Theodosius II. | Jobst 1977, pp. 41 and 52: a coin of Valentinian I (364–375) was found with a coin of Constantine beneath the original mosaic floor of the upper church. Strocka 1977, pp. 58–63, dates the decoration to the same period independently. Ladstätter and Magdalino 2019, p. 261, shows the shrine inserted secondarily into the standing fourth-century building, its east wall still carrying both the original rectangular niche and the later apse; at p. 41, as at Strocka p. 58, the site is a disused quarry used for Christian burial before Constantine. | |
| Ancient city | “… a 3rd-century earthquake, in the third quarter of the century, ended its life as housing, and parts of it then held workshops and mills.” | “An earthquake in the third quarter of the 3rd century ended its life as housing, and the excavators now exclude any return to living here: the ruin went over to metalworking, and in the 7th century to water mills.” The page was right, and is now more exact. | For Terrace House 2, continued residential use in late antiquity is ruled out in Sabine Ladstätter’s account in the excavation reports (Forschungen in Ephesos VIII/6, p. 265; Pülz 2020, pp. 300–302; Ladstätter and Magdalino 2019, p. 40), which record metalworking shops in the fifth and sixth centuries and a stone saw and water mills in the seventh. The comfortable late-antique apartments described by Foss belong to Terrace House 1, a separate block with its own story. | |
| Basilica of St John | “The Greek archaeologist Georgios Soteriou dug here in 1920–21, during the Greek occupation, and the Austrian Archaeological Institute excavated from 1926 to 1931, with the results published by Hans Hörmann in 1951.” | Both dates are corrected by a year. Soteriou began in 1921 and abandoned the work the following year; his own report appeared in the Athens Archaeological Society’s Deltion for 1921–22. The Austrian Archaeological Institute is usually credited with 1926–1931, but 1926 is the year Austrian fieldwork resumed at Ephesus as a whole — work on the church itself ran from the autumn of 1927 to 1931. | Hörmann 1951, p. 16, sets out the seasons: the narthex and the central burial installation in autumn 1927, the nave and deep soundings in 1928, the atrium and south arm by 1930, and 1931 for tidying up; p. 275 cites Soteriou’s own publication as Deltion VII, 1921–22. Elliger 1985, p. 195, has the church excavated since 1921, and Karydis 2016, p. 3, gives 1921 to 1922 and the Austrians from 1927. Ladstätter and Magdalino 2019, p. 14, dates the general resumption to 1926. | |
| Basilica of St John | The page did not say where the misread carved slab ended up. | A new paragraph follows the stone. The carved sarcophagus box was prised from the wall and acquired for the Duke of Bedford by 1828, and it is at Woburn Abbey in England today; a large fragment of its base was left behind, where Austrian excavators found it and Otto Benndorf photographed it in 1895 — both the fragment and the photograph have since been lost. | Kintrup 2017, pp. 49, 50, 60 and 75, in the catalogue of these objects: Spon and Wheler reported the reliefs in 1678, the box had reached the Duke of Bedford before 1828, Arundell sent the sixth Duke a further fragment in 1832 that turned out not to belong, and a related sea-thiasos fragment reached Vienna in 1911. Pülz 2010, p. 124, independently places the slab at Woburn Abbey. | |
| Basilica of St John | “… 18th-century travellers took them for scenes of Christian martyrdom. The name stuck.” | The century is no longer stated. The page reads that at some point early travellers took the reused Roman reliefs for scenes of Christian martyrdom, and then names both modern positions: Andreas Pülz, in the church’s own excavation series, calls the name modern and dates the mistake to the nineteenth century, while Carola Kintrup, in the study devoted to these sarcophagi, puts it with seventeenth- and eighteenth-century travellers. | Pülz 2010, p. 124, against Kintrup 2017, p. 170. Hörmann 1951, p. 17, and Elliger 1985, p. 205, describe the misreading but give it no century. The reliefs were on the wall and being described by 1678, when Spon and Wheler reported them (Kintrup 2017, pp. 49–50), and Richard Chandler, visiting in the eighteenth century, read them correctly as Achilles and Patroclus — so the mistake was never universal. There is one modern voice each way and no tie-break. | |
| Basilica of St John | “The Turkish Ministry of Culture gives the church six domes over a three-aisled, cross-shaped plan, while some English-language sources say five. The forecourt, or atrium, is given as about 47 by 34 m and rests on vaulted substructures, because the hillside falls away steeply.” | The bare disagreement is replaced by Karydis’s explanation: the ruin as excavated has six domed bays — two over the nave, three over the transept, one over the chancel — while Procopius compared the church to the five-domed Holy Apostles in Constantinople because he was describing the intermediate building of about 520, before the nave’s two bays were added. Five is that church; six is what visitors walk through. The atrium sentence stands, and a sentence on the Kantharos fountain is added. | Karydis 2016, p. 18, for the six bays visible in the excavated remains, and pp. 22–23 for the sequence that accounts for Procopius’s five. Procopius’s own text is known here at second hand, through Karydis. The atrium figure of about 47 by 34 m rests on German Wikipedia alone; the vaulted substructures are corroborated, the hillside falling away steeply behind them. | |
| Basilica of St John | “Justinian did not build on an empty site. He replaced a small, ruined older church … and the new building went up around the middle of the 6th century. Procopius records the rebuilding … The Austrian archaeologist Peter Scherrer names Theodora as co-builder with her husband, and capitals in the nave still carry the monograms of Justinian and the empress.” | The traditional single campaign is kept and attributed to Peter Scherrer, and Nikolaos Karydis’s two-campaign argument is then set out in his name: a fully vaulted, five-domed church on a Greek-cross plan standing by about 520, and a two-bay nave added around 548–550 — the only part whose capitals carry Theodora’s monogram as well as Justinian’s. She died in 548. | Karydis 2016 re-reads Hans Hörmann’s own excavation drawings: the west cross-arm’s foundations come in two thicknesses, 1.20 m against 0.70 m, which a single campaign never explained (p. 11), with the two phases at pp. 20 and 22 and the date of about 520 at p. 24. Foss 1979, p. 88, independently reads the nave monograms as showing that part finished before Theodora died, and both cite John of Ephesus on the ordination of seventy priests in the church’s precinct in 541. Procopius, Buildings V.i.4–7, is known here only at second hand. | |
| Guidebook errors | The architect was named simply as “an architect from Damascus, Ali ibn Mushaimish”. | The fuller form is used: Ali ibn Mushaimish al-Dimashqi, with a note in the copy that the last part of the name was taken from Damascus. The entry also records that naming the architect at all is unusual for a fourteenth-century Anatolian building. | Karabacek, in Forschungen in Ephesos I, p. 131, and Bayat, as reported by Ladstätter and Magdalino 2019, pp. 203 and 207, both print the shorter Ali ibn al-Dimashqi; the Turkish Religious Foundation’s encyclopedia of Islam gives the fuller Ali b. Müşeymeş ed-Dımaşkī. These are not rival readings: al-Dimashqi is the name taken from his city, Müşeymeş the patronymic the two European editions abbreviate away. Whether Müşeymeş stands on the stone, or comes from a later Ottoman source, is unresolved. | |
| Around Ephesus | The mosque was dated to March 1375. | It is dated to 13 March 1375, the more widely followed of two readings of the damaged month name on the inscription over the west door; the other gives January. The 30 January still met in print answers to neither reading and is dropped. | Ladstätter and Magdalino 2019, p. 207, following Bayat’s 1991 edition of the Arabic text, reads 9 Shawwal 776 AH — 13 March 1375. Karabacek, deciphering a plaster cast in 1906 for Forschungen in Ephesos I, p. 131, read 9 Shaʿban 776 AH, giving Saturday 13 January 1375, the weekday he himself printed. The readings fall exactly fifty-nine days apart, so what is in question is the month name on the stone, not the calendar. Both dates are Julian. Foss 1979, p. 158, gives 30 January. | |
| Ancient city | “Its builders, Mazaeus and Mithridates, were freedmen: former slaves of Augustus and of Agrippa respectively.” | “Mazaeus and Mithridates are often introduced as freedmen of two different men, Mazaeus of Augustus and Mithridates of Agrippa. Only half of that is evidenced: a tombstone names Mithridates as Agrippa’s freedman, while Mazaeus’s patron appears on no inscription and the excavation reports go opposite ways about him.” | A tombstone near the theatre, published in Forschungen in Ephesos III, p. 111, names Mithridates as Agrippa’s freedman, and the same report infers that Mazaeus shared his standing. Waldner 2020, p. 181, calls both men freedmen of Augustus. The gate’s own bilingual inscription honours Augustus, Livia, Agrippa and Julia and, in the Greek, addresses the dedicators’ former masters without naming either patron (Forschungen in Ephesos III, p. 64). Both were freedmen of the imperial household; which household stays open. | |
| Ancient city | “Sources give its length as anything from just over 500 m to 570 m.” | “Published lengths form two camps rather than a range, one around 500 m and one around 600 m, most likely because they take the street to begin and end in different places.” The roadway of about 11 m and the porticoes of some 5 m a side are unchanged. | For about 600 m: Elliger 1985, p. 179; the 1905 excavation report, Forschungen in Ephesos I, p. 64; and Dey 2015, p. 91. For about 500 m: Pülz 2010, p. 132, and Ladstätter and Magdalino 2019, p. 36. Two scholarly voices on each side, so neither figure is a stray, and the earlier range taken from guide sources is dropped. | |
| Ancient city | “It is roughly square, about 111 m a side, with three gates and around a hundred shops, workshops and storerooms behind its colonnades.” | “Two sizes are quoted and both are right: the courtyard inside the colonnades measures about 112 m each way, while the block as a whole, halls and the hundred or so shops behind them included, runs about 154 m a side.” Saying which figure is which removes a contradiction that never existed. | Rogers 1991, p. 100, and the two excavation reports — Forschungen in Ephesos III, p. 372, and Scherrer and Trinkl 2006, p. 372 — give the open inner courtyard as 111.89 by 111.39 m and the whole enclosed complex as about 154 m a side, some 24,000 m². Rogers 2012 quotes both figures together. The 110 m sometimes met is the same courtyard measurement, rounded. | |
| Ancient city | “Along its north side ran the Basilike Stoa, a hall about 167 m long dedicated in AD 11 by Gaius Sextilius Pollio …” — one length, given without qualification. | The dedication is stated as secure and the length given as two readings: about 167 m as usually published, and 195 by 20 m in Stefan Groh’s survey once the projecting end chambers are counted in. The figures are most likely measuring different extents rather than disagreeing. | Groh 2008, pp. 23 and 31, gives 195 by 20 m including the chalkidika, and uses the hall as the model for the Olympieion’s own south hall, later the Church of Mary, a third larger again at 265 by 30 m. No source with a page of its own supports the 167 m or the 160 m sometimes quoted. The AD 11 dedication by Pollio and his family to Artemis, Augustus and Tiberius, and the bulls’-head Ionic capitals, are not in question. | |
| Ancient city | “It was found in 1904, and in 1929 Josef Keil and Max Theuer opened the burial chamber and found a complete skeleton in a sarcophagus full of water.” | “Two dates attached to it disagree: the clearance is 1904 in the older literature and 1905 in the excavation volume, and the burial chamber, opened from above by Max Theuer with Josef Keil, is dated to 1929 in some accounts and 1926 in others.” Both readings are named rather than one chosen. | Waldner 2020, p. 89, credits the clearance to Rudolf Heberdey in 1905 and puts Theuer’s opening of the chamber in 1926; Hilke Thür gives 1926 as well, while the older literature has 1904 and 1929. The chamber’s other entrance, at the back, was reached only in 1982–83 under Hermann Vetters (Quatember 2017, p. 27), and Thür and Kanz re-inspected the monument in 2007 (Waldner 2020, p. 90). Both contradictions rest on one modern volume, so both readings stand. | |
| Great Theatre | The prose named no dissenting scholarly view on the start date beyond the popular and unsupported figure of 300–290 BC, which left the excavators’ second quarter of the second century BC as the only date with any standing. | A new paragraph names two further positions. Stefan Karwiese would date the built theatre only from about 100 BC, pointing to a well house at the north-west corner of the stage building as the earliest tangible construction evidence. Volker Michael Strocka dates a small fountain house at the same corner to about 280 BC — but it is a subsidiary building, not the auditorium, and its find context was disturbed by an earlier restoration, so it does not reopen the question of Lysimachus. | Krinzinger and Ruggendorfer 2017, p. 515, for the excavators’ own dating, with the Hellenistic orchestra at 12.35 m and a mid-second-century parallel at Metropolis; p. 513 for Strocka’s fountain house, dated by comparing its sima with the later Temple of Artemis. Karwiese’s c. 100 BC is reported through Rogers 2012, p. 344, and so is named rather than cited to a page of his own. | |
| Library of Celsus | In late antiquity the roofless hall served as the courtyard of a neighbouring house, and around AD 400 the front was turned into a fountain, with water running over the steps and the standing façade used as a display wall — given as the settled account. | The section presents an open dispute. The library’s own excavation report has the façade kept deliberately standing as the fountain’s backdrop; the later report on the neighbouring Terrace House has it already collapsed to its base course in the third-century destructions, so that the library stood as a ruin for a long generation before Stephanus’s basin went up in front of it. Neither account is picked. | Keil 1953, pp. 80 and 83, for the standing façade, with Stephanus’s basin faced in relief slabs from the Parthian Monument; Rathmayr 2016, p. 6, for the collapsed one — Hadrian’s Gate, the Temple of Hadrian and Terrace House 1 were repaired after the destructions of the 260s and 270s, and the library was not. The two reports cannot settle it between them; the answer lies in the façade’s own blocks. | |
| Library of Celsus | “Ennoia’s base names not Celsus but an otherwise unknown Philippos, in a painted rather than carved inscription, perhaps honouring a man who paid for that later repair.” | The “perhaps a repair donor” framing is replaced. Both modern readings agree that an existing honour was taken over, not added to: Josef Keil reads Philippos as the architect of the fountain conversion, rewarded with a re-dedicated statue, while James Harrison reads a plainer appropriation, Celsus’s own dedication for Ennoia removed to make room for Philippos’s. Who Philippos was remains open; that the base was re-labelled does not. | Keil 1953, p. 83; Harrison, in Harrison and Welborn 2018, p. 21, citing IEph 7.2.5110. A comparable inscription near the library, in which one Demeas replaces an Artemis with a cross, is dated about AD 435 on its letter forms and gives a rough benchmark (Keil 1953, p. 71). A further identification with Flavius Philippus, praetorian prefect of the East, who died in 351, has been suggested but is not followed here. | |
| Ancient city | “Its ancient name was probably Embolos; the old gloss ‘the wedge’ is now disputed — Foss rejects that etymology — and the Curetes Street page sets out the argument.” | “Its ancient name was probably Embolos, the ordinary late-antique word for a colonnaded street rather than the ‘wedge’ of the guidebooks; the modern one comes from lists of Curetes priests reused along it.” The gloss is gone rather than qualified, and the fuller argument stays on the Curetes Street page. | Foss 1979, p. 66, attributes the “wedge” reading to Hermann Vetters and rejects it: embolos was simply the ordinary late-antique word for a colonnaded street, and was used of other streets too. Foss 1979, p. 65, and Rogers 1991, p. 95, cover the Domitianic paving dedication and the late-antique verse inscription. German Wikipedia offers Plateia as a further possibility. | |
| Ancient city | “In the late Roman period its curved eastern end, about 50 by 42 m, was closed off as an arena for gladiators and wild-beast shows.” | “Its curved eastern end is usually described as closed off in the late Roman period to make an arena for gladiators and beast shows; Groh reads the same work as a small amphitheatre inserted into the curve, its seating bank roughly 40 by 50 m, shortening the track from about 232 m to 184.” | Groh 2008, p. 22, describes the inserted amphitheatre and sets it in an eastern pattern: amphitheatres appear late in the Greek east, and mostly in stadia that had lost their original use, as at Aphrodisias. Ladstätter and Magdalino 2019, p. 25, after Roueché, call the same almost circular installation an arena for beast hunts and gladiators; Krinzinger and Ruggendorfer 2017, p. 344, note that the stadium, not a purpose-built amphitheatre, was the usual venue in the Greek east. Arena or amphitheatre is a question of words, not of two buildings. | |
| Ancient city | The page said nothing about the hall at the east front, and the site’s own notes called it a ceremonial hall, the Kaisersaal, with a statue of Antoninus Pius in it. | “… whose statue stood in the marble hall at the east front. That hall has been called the Kaisersaal since the 1920s, a name Josef Keil coined and disliked; the excavation report prefers Marmorsaal, after the nine panels of patterned marble in its floor.” | Steskal and La Torre 2008, p. 4, records Keil’s own misgivings about the name, and p. 20 the nine fields of patterned marble that give the Marmorsaal its name. No inscription, relief, imperial portrait or fixing hole ties the room to emperor worship, and the niche block Keil read as a cult altar is doubted. The statue of Antoninus Pius is real; the cult hall is a label. The area of about 13,000 m² and the count of four bath-gymnasia are confirmed at pp. 1–2. | |
| Ancient city | “Whether it was dedicated to Artemis or to Aphrodite, together with Antoninus Pius, is disputed in the sources.” | “Its dedication has been reported two ways, to Artemis in some accounts and to Aphrodite in others. The building’s own inscription settles it: the goddess is Artemis, alongside Antoninus Pius.” | Steskal and La Torre 2008, pp. 1, 2 and 4, reading the building’s dedicatory inscription, with IvE 431 and 438, which also fix the dedication to AD 147–149 and name the donor’s wife Flavia Papiane as co-donor. Aphrodite is a guidebook error. | |
| Ancient city | “Once the basin below the city had been dredged into shape, the advancing delta forced the Romans to cut a long canal westwards to reach open water. The basin itself survives as a small marshy lake” — which left the harbour looking as though it had been lost in late antiquity. | “The basin was never deep, and it was not lost in one generation. Dredged repeatedly, latterly serving small craft, it stayed in use until the 11th or 12th century, and survives as a marshy lake.” The end of the harbour is no longer attributed to the eruption of Santorini in 726. | Ladstätter and Magdalino 2019, p. 122 — intensive use into the early seventh century, a later retaining wall, and no disuse before the eleventh or twelfth; p. 62, the shrunken basin still navigable for small boats in the tenth and eleventh, by which time the main anchorage had moved to the bay of Pamucak; p. 119, dredging from the first century to the late third, with Cyril of Alexandria arriving by barque in 431. Foss 1979, p. 186, disputes the 726 date on textual grounds. | |
| Temple of Hadrian | “Later generations named those baths after Scholastikia, the Christian woman who paid to renovate them in the 4th or 5th century” — a single loose span, given without qualification. | The usual phrasing is shown to hide a disagreement of more than a hundred years. Peter Scherrer puts her work around AD 400, Martin Steskal argues for the first half of the sixth century, and Alice Waldner ties the renovation to the same early-sixth-century programme that saw the inscribed Curetes columns carried down from the Prytaneion and reused along the street. Ursula Quatember holds that neither date can be proved on the present evidence. The inscription naming Scholastikia is secure; her date is not. | Quatember 2017, p. 130, sets Scherrer’s c. AD 400 against Steskal’s first half of the sixth century and declines to choose; Waldner 2020, p. 162, and Steskal 2010, p. 282, carry the early-sixth-century programme. Quatember 2017, p. 125, records the building’s names: the Kuretentherme, then the Scholastikiatherme once her statue was found, and only in 1978, on a newly published inscription, the Varius Baths — a name until then attached to a different building near the Upper Agora. | |
| Ancient city | “It seated about 1,400 people, or 1,500 in some accounts.” | “Counting the surviving rows gives over 1,600 seats, against the 1,400 or 1,500 of the encyclopedias.” | Bier 2011, p. 48, counts fifteen rows in five wedges in the lower cavea and ten rows in ten sections in the upper, inside a stilted semicircular retaining wall about 47 m across (p. 31). The 1,400 and 1,500 figures are encyclopedia estimates, and a much lower figure of about 300 also circulates without support. | |
| Ancient city | “It was built around AD 150 by Publius Vedius Antoninus and probably had a roof.” | “Publius Vedius Antoninus has long been called its founder, around AD 150; an inscribed letter found beneath the marble veneer he paid for shows a council house already standing by AD 128/129, so his gift was a rebuilding.” The Odeon page carries the fuller account. | Lionel Bier’s study of the building (Bier 2011, pp. 31, 54 and 81) found the inscribed letter under Vedius’s own veneer, which puts a council house on the site by 128/129 at the latest. Stefan Groh’s rival date of about 11 BC rests on survey rather than on the building itself, and its measurements do not fit the retaining wall of some 47 m, so it is not followed here. | |
| Ancient city | “They were repaired in the 4th century; the popular name ‘Baths of Constantine’ and its attribution to a particular emperor are not reliable.” | The attribution holds, once the right emperor is put into it. The work was done under Constantius II by Lucius Caelius Montius, proconsul of Asia, and the complex took the name Thermae Constantianae. The popular name garbles Constantius; it is not a mistake about Constantine. | Hanslmayr 2016, p. 19, and Steskal and La Torre 2008, p. 2, reach this independently, and Foss 1979, p. 60, had already read it from the inscriptions IvE 1314 and 1315. The regnal dates quoted in guidebooks are Constantius II’s; Constantine II ruled only from 337 to 340. | |
| Library of Celsus | “The monument itself is of the Antonine era, the mid-2nd century AD; older readings tie it specifically to Lucius Verus’s Parthian War of 161–166.” | The reliefs are still said to have been long tied to that war, with a date around 170 still quoted, but Alice Landskron’s case for AD 138 is now set out in her name. The battle frieze shows no identifiable Parthian campaign; the emperor plate — Hadrian, Antoninus Pius, Marcus Aurelius and a boy Lucius Verus at a sacrifice — can only show the succession settlement of the first half of 138, since Verus has no portrait type before 161. The readings of 161 to 172 are named as the older position. | Landskron 2008, pp. 62–70. The bearded head that Wilhelm Oberleitner read as a posthumous, veiled Lucius Verus carries no veil at all, which removes the main prop of the late date. Bier 2011, p. 98, had already argued for the early 140s from a different direction. Landskron is a single specialist voice, so the page names her and does not present AD 138 as settled. | |
| Temple of Artemis | “Franz Miltner’s Austrian team found them in 1956 in the Prytaneion, the city’s town hall: the Colossal or Great Artemis, the Beautiful Artemis and a smaller one” — one date for two different events, and a count of three given by listing. | The two events are separated. The Prytaneion was found in September 1955; the statues came out a year later, in a fortnight between 18 September and 3 October 1956 — the Beautiful Artemis first, then the smaller figure, then the Great. No count is printed: accounts give two, three or four, depending on whether a duplicate’s fragment is counted. | Steskal 2008, p. 364, gives the fortnight of 18 September to 3 October 1956 with the Selçuk inventory numbers 718, 712 and 717. Steskal 2010, pp. 56 and 58, gives 21 September 1955 for the start of the dig and the four find dates, including a duplicate’s foot fragment on 13 October. The count is genuinely disputed: German Wikipedia gives three, turkisharchaeonews two, and Rogers 2012, p. 180, four. | |
| Ancient city | “The building visible today is Augustan, was renovated in the Severan period and had been demolished before about AD 400.” A sentence under the Baths of Varius added that architectural pieces from the demolished Prytaneion were built into them, which showed it had come down by about 400. | “Miltner believed Christians had pulled the Augustan building down; the dedicated re-study finds no deliberate demolition, only damage from the earthquakes of about AD 358–368 that was never repaired.” The sentence about the baths, whose inference rested on the superseded demolition date, has been removed in the same pass. | Martin Steskal’s study of the building (Steskal 2010, pp. 128 and 281) finds no burning, no smashed fittings and no sign of deliberate demolition; stone was carted off in earnest only in the first half of the sixth century, and a coin shows people about the site into the last quarter of the eighth. Steskal 2008, p. 364, covers the find circumstances. Miltner’s Christian-destruction reading, and his story that Scholastikia toppled the great Artemis, are superseded. | |
| History | “Austrian excavations south of the Church of Mary found people living there until the 12th century, then mainly farming, and the last signs of any activity date from the 14th century.” | Both the old ninth- or tenth-century abandonment and this page’s twelfth century are named as overshooting. The lower city was still lived in through the eleventh century and still drawing Christian pilgrims, with the sanctuary of the Church of Mary rebuilt as a templon in that century. Only after the eleventh do the finds run out, leaving dry-stone rings read as animal pens or garden borders; burial around the church carried on into the fourteenth. | Ladstätter and Magdalino 2019, pp. 61 and 63, for the eleventh-century occupation, the pilgrim traffic and the disappearance of the finds afterwards; Degasperi 2013, p. 97, for the burials to at least the fourteenth century and for building work ending after the eleventh. The earlier ninth- or tenth-century date rested on an abstract of the very work that argues against it. | |
| Temple of Hadrian | “In front of the porch stand four inscribed bases. They carried statues of the First Tetrarchy, the four emperors who ruled jointly: Diocletian, Maximian, Constantius Chlorus and Galerius. Maximian’s statue was later replaced with one of the father of Theodosius I.” | Four statues stood along the street front, but they did not arrive together and the bases are not one donation. Iunius Tiberianus, proconsul of Asia, gave the bases for three Tetrarchs between 293 and 303 — Galerius, Diocletian and Constantius I, west to east. Maximian’s statue stood before the western column until his memory was condemned, and between 379 and 387 the proconsul Nummius Aemilianus Dexter replaced it with one of the father of Theodosius I. | Quatember 2017, pp. 84–85, with the inscriptions IvE 305 and 306; Foss 1979, p. 30, corroborates the Theodosius dedication. The Constantine of the guidebooks goes back to the excavators’ field diary of 17 October 1957, and the building study identifies that ruler as Constantius I (Quatember 2017, p. 378). Whether Maximian’s base belonged to the same donation as the other three is unresolved, which is why the page counts four statues rather than four bases. | |
| Temple of Hadrian | “The carved frieze inside the porch is not Hadrianic. It was carved in the 3rd or 4th century, and some of its slabs probably came from a different monument, reused here after a 4th-century earthquake.” The fourth slab was described as often identified with Theodosius I and his family, that reading being debated. | The section is rewritten as an explained dispute. Miltner’s Theodosian date, Fleischer’s third quarter of the fourth century and Brenk’s Tetrarchic reading are given as the older positions. Then Ursula Quatember’s building analysis: the porch’s north-east corner was never structurally altered, so the relief blocks bonded into it were never lifted, and the frieze belongs with the building, in the first quarter of the second century. Only the west side of the porch collapsed and was rebuilt. Fleischer’s reading of the scenes stands; his date and the reuse clause do not. | Quatember 2017, p. 82 and pp. 105–111, with p. 83 for the fourth-century rebuilding of the west porch and p. 22 for the three older positions. Peter Scherrer reached the same conclusion independently, comparing the carving with the Arch of Titus, the Arch of Trajan at Benevento and the Hadrianeum in Rome (reported at p. 110). Miltner read the fourth block as Theodosius I’s family, which cannot stand if the blocks are original; Scherrer proposes the recently dead Trajan worked into the assembly of gods. Dionysus, Cybele and an unnamed emperor remain open. | |
| Ancient city | “Around 614–616 large parts were probably destroyed again, and the wall built afterwards took in less than half of the ancient city.” | The date is now anchored in coins rather than chronicles. Destruction layers across the city carry coins of Phocas and the early years of Heraclius, putting them at about 610 to 616, some into the 620s. The cause stays open — the Persian war or an earthquake — as does whether this was one catastrophe or several. The wall that followed was standing by 609/610 and enclosed less than half the old city. | Scherrer 2007, p. 339, where almost every known urban context ends in a destruction layer with coins of Phocas or the early Heraclius; Scherrer and Trinkl 2006, p. 54, finds the same break in every quarter from the Magnesian Gate to the market, sometimes with burning. Krinzinger and Ruggendorfer 2017, p. 119, dates the wall before 609/610 from imperial acclamations carved at the gate west of the theatre. Ladstätter and Magdalino 2019, p. 37, doubts the single-catastrophe model. | |
| Temple of Artemis | “Modern scholars reject that reading or call it unproven, and offer ornaments hung on the wooden statue, the scrota of sacrificed bulls, eggs, amber pendants or simply regalia in its place, none of which has won general agreement.” | The qualification stands, but the dispute is named rather than summarised away. Gerhard Seiterle’s bull-scrota proposal is reported seriously by Dieter Knibbe and leaned towards by Guy MacLean Rogers; Rick Strelan rejects it because the analogy rests on a story about a quite different goddess at Autun, and Lynn LiDonnici finds it poorly supported. Strelan’s own suggestion, that the shapes resemble the trunk of a date palm below its fronds, is added. | Strelan 1996, p. 89, for the rejection, the date-palm reading and LiDonnici’s verdict; Knibbe, in Koester 1995, p. 142, reports the theory and the questioning; Rogers 2012, p. 357, wonders whether it points to the mystic bull sacrifice Strabo alludes to. No ancient source describes the objects at all, which was Robert Fleischer’s original objection to Seiterle, and is why the question stays open. | |
| Terrace Houses | “The houses were never rebuilt as grand residences, and the disasters of that century ended the city’s imperial heyday, a turning point in the history of Ephesus.” | “The houses were never rebuilt as grand residences; whether that run of disasters really ended the city’s imperial peak is disputed among historians, as the history of Ephesus sets out.” The first half stands; the second is presented as contested, with the fuller treatment on the history page. | White, in Koester 1995, p. 29, holds that Ephesus came through the third- and fourth-century transition that finished other cities of the east; Pülz 2020, p. 323, describes a busy metropolis again from the end of the fourth century, its life gathered around the harbour gymnasium and the great halls beside it. What ended in the third century was one kind of aristocratic display, not the city. | |
| History | The third-century damage was told as a single sequence: an earthquake under Gallienus, Gothic raids from the sea, and Terrace House 2 destroyed in the third quarter of the century. | The dates are unchanged. A sentence has been added naming Sabine Ladstätter’s doubt about the single-catastrophe model: she finds some fifty years between the destruction along Curetes Street and the destruction south of the Church of Mary, and prefers several episodes to one ruinous year. | Scherrer 2007, p. 339, for the earthquake under Gallienus and the plundering that followed; Ladstätter and Magdalino 2019, p. 37, for the doubt. The year 262 comes from the Historia Augusta and is traditional rather than archaeological, while excavation dates the fall of Terrace House 2 to the third quarter of the century. This is a different event from the later, Heraclian destruction described in the section on the end of the city. | |
| Where is Ephesus? | “Tradition dates that settlement to the 11th or 10th century BC; the oldest Iron Age Greek pottery dug up at the site belongs to the late 11th.” | The dates stand, but the pottery clause is tightened. The early pottery comes from the sanctuary terracing under the Artemision, not from a settlement: no town of that date has yet been found. The page now keeps the sanctuary and the town apart. | Elliger 1985, p. 18, puts the main Ionian migration between 1100 and 800 BC, a wider window than the eleventh or tenth century; at p. 20 he records that no early Ionian finds had been made at Ephesus itself, which is why he argues from Old Smyrna instead. Scherrer 2007, p. 327, finds nothing on the western, coastal side earlier than the mid-eighth century. Klebinder-Gauß 2007, p. 197, records Protogeometric pottery in a fill layer at the Artemision. | |
| History | The capital question was qualified but named nobody: “Whether it was formally the capital … is debated.” | The two works that state it outright are named — Winfried Elliger, for whom Ephesus replaced Pergamon under Augustus, and Fredrick Long, who dates the change to 29 BC — and the reason for keeping the qualification is given. The province had no capital in the modern sense: the proconsul travelled a circuit of assize towns, and what the law required was that he make his landfall at Ephesus. “First and greatest metropolis of Asia” is identified as a civic title, not an administrative rank. | Elliger 1985, p. 61; Long, in Harrison and Welborn 2018, p. 200. Against a constitutional capital: Ulpian, Digest 1.16.4.5–6, reported by Rogers 2012, p. 350, which requires the landfall and nothing more; and Forschungen in Ephesos II/1, p. 379, where the city’s own inscriptions style it first and greatest metropolis. No ancient text constitutes Ephesus as caput provinciae. | |
| History | “He also destroyed the nearby cities of Lebedos and Colophon and moved their people into his new Ephesus” — asserted in the page’s own voice. | The claim is attributed to Pausanias and then qualified. Both cities are still there afterwards: Lysimachus’s own officer secures Colophon a few years later, and Lebedians appear at Ephesus as a named division of the citizen body rather than as refugees. The forced transfer of people is credible; the levelling of two cities is not. | Elliger 1985, p. 48, calls the destruction an exaggeration; Rogers 2012, p. 67, reports it rather than asserting it; Rogers 1991, p. 99, describes only forcible relocation. Rogers 2012, p. 43, has Prepelaos securing Teos and Colophon in 302 BC, and p. 344 has settlers from Lebedos forming the fourth chiliastys of the Ephesians. The ancient source is Pausanias 1.9.7 and 7.3.4–5, and the modern accounts are unanimous in refusing to take it straight. | |
| Ancient city | “Fieldwork stopped for the two world wars, resumed in 1926 and restarted with annual campaigns under Franz Miltner in 1956.” | “Fieldwork stopped for the two world wars, resumed in 1926 and, with Franz Miltner as director from 1954, returned to annual campaigns in 1956.” The two years were never a disagreement but two different facts: Miltner became director in 1954, and annual work resumed in 1956 after a gap of nineteen years. | The Austrian Archaeological Institute’s own account of its research history gives the sequence — interrupted by the First World War for over a decade, resumed in 1926 on the bath-gymnasium complexes and the Christian monuments, halted again by the Second World War, and back to annual campaigns in 1956. Ladstätter and Magdalino 2019, p. 15, has Miltner as director of excavations from 1954. He died on 23 July 1959. | |
| History | “Croesus took Ephesus around 560 BC” — a conquest, stated without qualification. | Herodotus’s fuller account comes first: the Ephesians were the first Greeks Croesus attacked, and by the end of the campaign the Greeks of Asia were paying him tribute. Then the modern readings, each named — Winfried Elliger has the city left standing under tribute with the unpopular Pindaros sent away, Guy MacLean Rogers has Pindaros expelled outright, Christine Thomas has Croesus spare the city altogether. The page adds that no ancient source gives a year: “about 560 BC” is taken from Croesus’s accession. | Herodotus 1.26–27, in Macaulay’s translation, supports conquest and tribute together, which is why the modern accounts can each be right about their own half. Elliger 1985, p. 31; Rogers 2012, p. 31; Thomas, in Koester 1995, p. 99, following Aelian, Varia Historia 3.26, and Polyaenus 6.50. Keil, in Forschungen in Ephesos I, p. 31, adds that only the Ayasuluk settlement could have been besieged as the city then lay. | |
| History | “after the revolt of Aristonicus was crushed, the province of Asia was organised in 129 BC”, with the summary table giving “organised 129 BC”. | 129 BC is named as the reference-book date and two specialists are given for 126: Steven Friesen, for whom Asia formally became a province in that year, and Guy MacLean Rogers, who has Manius Aquillius completing the organisation by then and dividing the province into assize districts, with Ephesus among the original court centres. The honest form is stated — constituted after 129 BC, organised by 126 — and the summary table now reads “129–126 BC”. | Friesen 1993, p. 3, and Rogers 1991, p. 3, against four encyclopedia entries for 129, which are one class of source rather than two. The question is epigraphic and prosopographic — when the provincial law of Aquillius and his commission of ten took effect — and nothing in the ancient evidence settles it, so the page gives the range rather than a verdict. | |
| Temple of Hadrian | “He probably came back in 128/129, and Ephesus honoured him as a founder” — given without qualification, inside a paragraph that read the imperial visits as the high point of the city’s Roman fortunes. | The visits and the founder honours stand, but the warmth of the relationship is now presented as argued. The page names Alice Landskron’s minority reading — that relations stayed cool, that Pergamon and Smyrna drew far more of Hadrian’s attention and money, and that the building boom belongs to Antoninus Pius — as one scholar’s case rather than a finding, and links to the history page, which sets it beside the evidence running the other way. | Landskron 2008, p. 76, is a single voice and is attributed by name rather than stated as fact. Against it: White, in Koester 1995, p. 54, where dated construction quickens under Trajan and Hadrian and runs on through the Antonines; and Sokolicek et al. 2025, p. 47, for the grain decree of about AD 129, the cleaning of the harbour and the diversion of the Kaystros. What is open is the character of the reign at Ephesus, not the visits. | |
| History | The foundation paragraph put Androclus’s settlers on “the slopes of Mount Koressos”, making Koressos one of the two hills above the city. | The page reads “the slopes of Koressos”. Koressos was not a mountain at Ephesus but a quarter of the city, with its own harbour and gate, on the far side of the bay near the Roman stadium. The two hills keep their own ancient names: Panayırdağ is Pion, Bülbüldağ is Preon or Lepre Akte. | Keil, in Forschungen in Ephesos III, p. 116: an inscription places a broad street “in Koressos” between the stadium and the Artemision, and the only Mount Koressos in an ancient text lies forty stadia off. Scherrer 2007, p. 327, reaches the same conclusion and notes that the distance from that quarter to the sanctuary is the seven stades Herodotus 1.26 gives for the rope the Ephesians ran from temple to wall. | |
| History | Strabo’s story was told as straight narrative — Lysimachus waiting for a downpour, blocking the drains and flooding the old town — and the geoarchaeological silting was then offered as “a more practical motive”, so that the two read as competing explanations. | The story is kept but framed as hostile and moralised. Peter Scherrer, following Dieter Knibbe, argues that the drains really did back up, because the rising sea had closed their outfall, so that a real and unglamorous problem was later remembered as a king’s cunning. Kraft and his colleagues reach the same conclusion from the delta sediments and dismiss the other stock motive, that the move was vanity in the service of the city’s new name. | Scherrer 2007, pp. 344 and 345; Kraft et al. 2007, p. 135; Rogers 2012, pp. 60 and 63, on the refoundation itself. Strabo 14.1.21 is the only ancient account of the ruse and was composed some three centuries after the event; nothing in it makes the episode more than a story. | |
| History | Croesus’s gathering-in of the population was given as Lydian politics, following Strabo’s sequence without offering any cause for it. | The page adds Peter Scherrer’s environmental reading and attributes it to him: by the mid-sixth century a rising shoreline had put the scattered harbours round the bay out of use, so the move to one city beside the sanctuary was practical as much as political, and Croesus’s gifts to Artemis gave a merged population of Greeks, Lydians and Carians something to hold in common. Stefan Groh’s caution comes with it — burials continue at several of the old sites, so the gathering-in was less tidy than the word synoikismos suggests. | Scherrer 2007, p. 340; Groh, in the Tetragonos Agora excavation report (Scherrer and Trinkl 2006, p. 267). Strabo 14.1.21 says only that the Ephesians lived on the mountainside until Croesus’s day and afterwards came down to live around the Artemision. Scherrer is one voice, and the page gives his reading as an argument rather than as a finding. | |
| History | No paragraph distinguished the two things the coast did. It described a shoreline that simply kept retreating west for some seven thousand years as the Kaystros pushed its delta out, which left the silting of the old harbour and Strabo’s story of a flooded town reading as rival explanations for the same move. | A new paragraph separates the phases. Through the Archaic and Hellenistic centuries the sea was rising, drowning quays and breakwaters and backing water up the drains; only from the Roman period does the delta advance west and strand each harbour in turn. The reading is attributed to Peter Scherrer, working from the geoarchaeology of John Kraft and his colleagues, and the early silting is said to have come off the hills to the south rather than from the great river. | Scherrer 2007, pp. 340 and 345, for the two phases and for the southern streams; Kraft et al. 2007, p. 135, independently reads Lysimachus as a planner responding to the advancing delta. Strabo 14.1.21 gives the sequence of settlements but no cause, which is the gap the geoarchaeology fills. | |
| Temple of Artemis | The page dismissed the whole episode: “The popular claim that Christians destroyed the temple in AD 401 is not well supported,” adding that Cyril of Alexandria credited John Chrysostom with overthrowing it but that there was little evidence for that either. The summary of common errors repeated the dismissal. | The paragraph sets out a real disagreement instead. Three modern specialists accept that Chrysostom’s visit of 401 ended the sanctuary, and near-contemporaries assume it, but none of them describes a demolition, and the building had lain ruined since the Gothic raid of the 260s — so 401 marks the end of the cult rather than of the temple. The summary now says that the ancient word is “despoiled”. | Foss 1979, p. 86; Limberis, in Koester 1995, p. 334; Rogers 2012, p. 253 — three independent voices for the despoliation of 401 (Karydis 2016 reports it from Foss). The older dismissal rested on encyclopedia sources alone. Among ancient texts, Palladius, Dialogue 14–15, describes the visit — a synod and six deposed bishops — without mentioning the Artemision; the overthrow language is Cyril, Homily XI, and Proclus, Oration XX.3, with Prudentius, Contra Symmachum II.495 (c. 402) and Paulinus of Nola, Poem XIX.95 (405) already assuming it. |