BMCR 2024.02.37

The painted Tetrarchic reliefs of Nicomedia: uncovering the colourful life of Diocletian’s forgotten capital

, The painted Tetrarchic reliefs of Nicomedia: uncovering the colourful life of Diocletian's forgotten capital. Studies in Classical Archaeology, 12. Turnhout: Brepols, 2021. Pp. xviii, 198. ISBN 9782503594781.

It is hard to overestimate the importance of this volume. The painted marble reliefs of Nicomedia, excavated in the Çukurbağ neighbourhood of İzmit, are the best-preserved examples of polychrome relief sculpture in Roman art, and they represent Nicomedia when it was arguably the most powerful city in the Roman empire, the favourite city of Diocletian. Nicomedia is hidden beneath modern İzmit, and the sixty-six relief panels (and other finds) were uncovered through rescue excavations in 2001 and 2009. In 2015-2018, the reliefs were studied and identified thanks to the TÜBİTAK Çukurbağ Archaeological Project, and the project’s director, Tuna Şare Ağtürk, has produced the primary publication.

Chapter 1 surveys the history and layout of Nicomedia and the archaeological research conducted in İzmit. It then presents the architectural context studied thus far, which shows that the reliefs decorated a two-storey aediculated imperial hall within a Dyarchic/Tetrarchic palace. Chapter 2 examines the style of the reliefs, the techniques employed, and the use of colour. The naturalistic reliefs were likely produced by well-attested Nicomedian workshops (p. 26), and the project was rushed. For example, finished parts of reliefs were shaved off during the fitting process (pp. 34-35), and the driplines of paint suggest that reliefs were painted in a vertical position after being fitted into their architectural setting (p. 36).

The paint was studied using in-situ microscopic examination, UV, IR and VIL imaging, and pXRF analysis. This allowed Şare Ağtürk to establish the prevalence of imperial purple, by which she identifies various figures as emperors. She demonstrates that painters employed colour to depict ostentatious imperial clothing and, in an aduentus scene, to indicate the greater age and seniority of Diocletian vis-à-vis Maximian (pp. 38-43). Painters also used polychromy to improve the legibility of visual elements when viewed from a distance. Notably, a group of otherwise identical barbarians are distinguished from one another by their differently coloured trousers (pp. 37-38). Paint often supplies additional details not present in the carving, which points to the close collaboration between sculptors and painters (pp. 36-37).

Chapter 3 examines the iconography of panels that depict the emperors or reference their achievements in war and peace. Şare Ağtürk notes that the reliefs never reference an imperial college of four. Moreover, some of the panels employ physiognomic features to distinguish between Diocletian and Maximian in a manner similar to certain coins and medallions of the Dyarchy and early First Tetrarchy, in contrast to the later emphasis on Tetrarchic similitudo (pp. 49, 55-57). For these reasons, Şare Ağtürk proposes that the reliefs date to the Dyarchy (285/6-293). However, the panels that survive do not represent the entire artistic program, and as discussed by Şare Ağtürk, the reliefs “exhibit a lingering Greek taste for naturalistic expression and Hellenistic mannerisms” (p. 27), as does the famous “head of Diocletian” found in İzmit’s Seka area (p. 28). This reviewer wonders whether regional taste better explains the approach to physiognomy, and whether the Caesars are truly absent. An emperor, depicted in combat against northern barbarians, uses a lion-skin saddle and has Maximian’s reddish-brown beard, but the nose is long and not the retroussé nose of Maximian (p. 110). I propose that this is the Herculean Caesar Constantius.[1]

Şare Ağtürk identifies battle scenes and ceremonial compositions, as well as references to the signa Jovius and Herculius. She provides a consistently lucid analysis and makes repeated use of other Tetrarchic monuments and media to strengthen her arguments. Şare Ağtürk identifies one panel where an emperor hands an incoming consul a codicil-diptych, and she notes that the relief “could be the earliest illustration of the type which became popular in the imagery of fourth- and fifth-century consular diptychs” (p. 65). Şare Ağtürk suggests that the panel commemorates the inauguration of a consul in Nicomedia, and she offers Cassius Dio (cos. 291) and Hannibalianus (cos. 292) as candidates. If one accepts her Dyarchic dating of the reliefs, Januarianus (cos. 288) is also a possibility.[2]

Şare Ağtürk persuasively identifies another panel as the forced migration of Laeti (first explicitly attested in 296/7), and discusses the possibility that this was yet another Tetrarchic innovation and thus a new element within triumphal imagery (pp. 60-63). She identifies the earliest iconographic instances of gods dragging enemies and demonstrates that the barbarians depicted across various panels consist of two distinct groups, including a long-bearded group wearing Phrygian caps. Noting the futility in attempting to identify specific ethnicities, Şare Ağtürk argues that the differentiation of barbarians advertises success broadly across the empire (pp. 65-69). I think this is fundamentally correct, but as Phrygian caps were associated with peoples in eastern Europe and the Middle East, viewers may have associated one group with victories in the east and the other with the west, thus their symmetrical representation on either side of Athena (p. 69).

Relief block with figures and horses in outdoor landscape
Migration of the Laeti. Image source: Mark B. Abbe and Tuna Şare Ağtürk, “The new corpus of painted Imperial Roman marble reliefs from Nicomedia: a preliminary report on polychromy,” Techne 48 (2019), 100-109. Open access:

Most notably, Şare Ağtürk provides a compelling look at the aduentus scene, a centrepiece within the artistic program, in which Diocletian and Maximian reunite with one another (pp. 54-59). She discusses the similarity in costume but also the seniority of Diocletian, as well as the typically late antique approaches to costumes, paraphernalia, and hierarchic scale. Having dated the reliefs to the Dyarchy, she suggests that the frieze represents the earliest known example of the Tetrarchic embrace and notes the possible significance of the fact that the emperors are not yet depicted frontally, as came to be typical in late antiquity.

Relief block with embracing emperors
Aduentus scene. Image source: Mark B. Abbe and Tuna Şare Ağtürk, “The new corpus of painted Imperial Roman marble reliefs from Nicomedia: a preliminary report on polychromy,” Techne 48 (2019), 100-109. Open access:

Piecing together the other fragments of the aduentus scene, Şare Ağtürk identifies two togate men, a togate boy, and the arm of another child as possible members of the imperial family (pp. 59, 124). As another reviewer, Anne Hunnell Chen, has pointed out, there is a procession of togati, including a boy, on the surviving base of the “Five-Column Monument,” erected in Rome in 303, a scene that is generally understood to represent the senatorial order.[3] Hunnell Chen is likely correct that the corresponding imagery at Nicomedia represents the clarissimi. Indeed, this may suggest that the embracing emperors are in Rome in 303, when they reunited for their uicennalia. However, Şare Ağtürk’s suggestion reminds this reviewer that certain nuances of Tetrarchic dynastic practice remain obscure. In many ways the Tetrarchs resisted dynastic norms. For example, they overlooked biological sons in planning the succession, and their coinage did not depict imperial women until 307.[4] However, before the Tetrarchic succession events of 293 and 305 temporarily changed the representational and ideological landscape, was there a time when imperial children enjoyed greater visibility within Dyarchic media? If Şare Ağtürk’s Dyarchic date is correct, it is relevant that a panegyric delivered in 289 honours the child Maxentius as if he were Maximian’s heir.[5] And were there architectural and figural contexts of visibility deemed more acceptable for family members who were not emperors? A panegyrist in 307, describing events during the First Tetrarchy, claims to have seen a painting of Constantine and Fausta in an imperial dining hall.[6] Most relevant to the togate imagery on the Nicomedia frieze, Maxentius and his son Romulus were clarissimi.[7] Within Nicomedia itself, Diocletian constructed palaces for his wife and daughter, and the “palace of Fausta” in Rome may have been constructed under Maximian.[8] In Romuliana, yet another palatial context, Galerius constructed a large tumulus to commemorate the deification of his mother, the palace’s occupant, and Maximinus appears to have followed Galerius’ example, building a tumulus for his mother at Šarkamen that likely also suggests deification.[9] I thus hope to see further discussion on these Tetrarchic togati.

In Chapter 4, Şare Ağtürk demonstrates that two panels from the Nicomedia frieze – one originally of unknown provenance – depict 1) the city’s foundation by Nicomedes I; and 2) a newly identified foundation myth connecting Nicomedia to Medea. Her interpretation is persuasive, as Nicomedia’s rivals in Asia Minor also manufactured alternative foundation myths to locate their heritage deep within mythical antiquity. Şare Ağtürk argues that the unusual depiction of a Fury alongside Medea casts the mythical founder in a more sympathetic light (p. 78).

Chapter 5 discusses a series of agonistic scenes, which, as an inscription indicates, relate to festivals of the imperial cult, another focus of competition between the cities of Asia Minor. The chapter includes a fragment depicting a wildcat and an elephant driven by a mahout. Şare Ağtürk suggests that the scene once depicted a procession of diplomatic gifts, and she relates it to a panel on the Arch of Galerius, where the Tetrarchs receive elephants and wildcats from the defeated Persians (pp. 85-86).[10] If one accepts a Dyarchic date, then I suggest the scene be linked to an episode in the 280s when Bahram II sent gifts, including eximiae pulchritudinis feras.[11] An elephant with a mahout, while alluding to Dionysus’ triumph, is especially fitting considering the Sasanian use of war elephants. Although the elephant’s ears and back possibly identify it as African (p. 148), it nonetheless resembles the example on the Arch of Galerius.

In Chapter 6, Şare Ağtürk synthesizes her findings to conclude that the Nicomedia frieze “celebrated not just the imperial cult of the co-emperors, but also the city of Nicomedia as the new seat of this imperial power” (p. 90). This is persuasive, and it may be illuminating to expand the theme of inter-urban competition beyond Asia Minor. For although Nicomedia does appear to have been Diocletian’s favourite sedes imperii,[12] both Sirmium and Antioch also served that function for him at various times.[13] The final chapter presents a detailed catalogue of the relief panels, and there is a catalogue of smaller relief fragments in the Appendix.

It is to the resounding credit of Şare Ağtürk that she has produced such a rich study, which has inspired observations and suggestions for future research. This volume showcases a truly exceptional and ground-breaking series of finds, but additionally, Şare Ağtürk has provided a well written and beautifully illustrated study, whose analyses have injected exciting new possibilities into research on Roman art, Asia Minor, late antique emperorship and the Tetrarchy. This volume will be a necessary reference for every one of these areas of study. I look forward to the publication of the architectural elements and free-standing statues also found in the Çukurbağ neighbourhood.



[1] Interestingly, this scene prominently features a soldier in scale armour, like the soldiers in the adlocutio scene on the Arch of Galerius in Thessalonica, and like one of the Caesars depicted in the crown of the “head of Galerius” from Romuliana.

[2] Timothy D. Barnes, The New Empire of Diocletian and Constantine (Harvard, 1982), 98.

[3] Anne Hunnell Chen, “Dyarchs in Color,” JRA 35.2 (2023), 997, who also argues against the suggestion that a togatus is being crowned by Victory.

[4] See e.g. Byron Waldron, Dynastic Politics in the Age of Diocletian, AD 284-311 (Edinburgh, 2022).

[5] Pan. Lat. X(2).14.1-2.

[6] Pan. Lat. VII(6).6. For contrasting views on the painting and the betrothal being depicted, see Waldron, Dynastic Politics, 176-177, 188-192; Julia Hillner, Helena Augusta: Mother of the Empire (Oxford, 2023), 65-68.

[7] ILS 666-667. Either the Tetrarchs denied biological sons the right to be nobilissimi, or Maxentius lost his nobilissimate status when his father abdicated, with the grandson’s status being patrilineal.

[8] Hillner, Helena Augusta, 60.

[9] Hillner, Helena Augusta, 82-96. In Dynastic Politics, 216-219, I dated Romula’s deification to 305/6, which is possible, but admittedly there is no positive evidence that the event post-dated Diocletian’s abdication. Coins depicting diuae were uncommon, which may suggest that female deification was unusual in the Roman empire, and Romula was a unique diua in that she was a mother with no imperial connections prior to Galerius’ co-option as Caesar. However, it is striking that Galerius went so far as to deify his mother and yet did not mint consecratio coins.

[10] B.I.18 in Margret S. Pond Rothman, “The Thematic Organisation of the Panel Reliefs on the Arch of Galerius,” AJA 81.4 (1977), 427-454. See also the “Matrix of Olbia,” Chron. Min. 1.148 and Lactant. Mort. Pers. 16.6 presenting elephants in Galerius’ triumph.

[11] Pan. Lat. X(2).10.6-7.

[12] Lactant. Mort. Pers. 7.8-10.

[13] Barnes, New Empire, 49-56.