Introduction

Reconstruction by Hebrard Ernest Hebrard's reconstruction of Split, Paris 1911

Split - or Spalato - is one of the most extraordinary places of the later Roman world, being no less than the palace which the Emperor Diocletian began building in 293 AD in readiness for his retirement from politics in 305. On the Dalmatian coast, adjacent to the Roman city of Salonae, it takes the dual form of a legionary camp similar to those still to be seen on the frontiers of Syria (appropriately so, for Diocletian was of necessity a military emperor) but also, with its splendid loggias, of an Italian house.

The name "Split" is, for the fanciful down the ages, a contraction of "Spalatum" - that is "palatium" or "palace". A similar example of folk etymology is to be found in Sicily, where the late Roman villa with prestigious mosaics, at Casale, is near a town called Piazza Armerina - "Piazza" here likely being derived from "palatium". Constantine Porphyrogenitus certainly thought so - but contemporary opinion holds that the name more likely derives from the Greek name for the area - Aspalathos, which is a shrub.

The importance of Split resides both in its state of preservation, and in the dearth of comparable examples from the Roman world. There are no coherent palace structures left in Italy, for example: fragments exist at Ravenna, althpough they are difficult to identify; the Palatine Hill in Rome (the origin of the word "palace", because that is where the Imperial palaces were) presents several overlapping structures - but nothing in so coherent a form as Split, where the structure of the palace/camp tells us much about imperial ceremonial and god-like pretensions.

Edward Gibbon gives a good description of how Diocletian, retiring (without being pushed) from the Imperial Purple, came here to grow cabbages:

A miserable village still preserves the name of Salona; but so late as the sixteenth century the remains of a theatre, and a confused prospect of broken arches and marble columns, continued to attest to its ancient splendour. About six or seven miles from the city Diocletian constructed a magnificent palace, and we may infer, from the greatness of the work, how long he had meditated his design of abdicating the empire ...

Though Constantine, from a very obvious prejudice, affects to mention the palace of Diocletian with contempt, yet one of their successors, who could only see it in a neglected and mutilated state, celebrates its magnificence in terms of the highest admiration ... The form was quadrangular, flanked by sixteen towers ... The whole was constructed of a beautiful free-stone, extracted from the neighbouring quarries of Trau, or Tragutium, and very little inferior to marble itself. Four streets, intersecting each other at right angles, divided the several parts of this great edifice, and the approach to the principal apartment was from a very stately entrance, which is still called the Golden Gate. The approach was terminated by a peristylium of granite columns, on the one side of which we discover the square temple of Aesculapius, on the other the octagon temple of Jupiter...

The range of principal apartments was protected towards the south-west by a portico five hundred and seventeen feet long, which must have formed a very noble and delightful walk, when the beauties of painting and sculpture were added to those of the prospect (Gibbon 1960, 135-6).

The ground-plan is a trapezoid, with the south (sea) side (157.5 metres) endowed with a splendid balcony but only a small gate. The longer walls are on the east (191.25m) and west (192.10m) sides and these, together with the closing wall to the north (150.9m) have impressive gates. The walls are some 17m in height, and 2m thick, and are largely intact, with square towers at the corners and on the long sides, and fancier octagonal ones flanking all the landward gates.

Palace or Chateau?

Although the site is traditionally called the Palace of Diocletian, and has monuments in a setting suitable for the enactment of imperial ceremonies, there is no universal agreement that - for example - the peristyle was so used, or indeed that Diocletian after his retirement was due for or received any cermonial attention.

The whole question is confused by our limited knowledge of just how ceremonial functioned at this time. The current state of play seems to be that the pro-palace lobby (Dyggve 1941, L'Orange 1965) has lost to the pro-villa lobby (Duval 1965, Wilkes 1986), who point to the achronicity of referring to ceremonial from the Byzantine court (i.e. several centuries later) and trying to fit it to a setup of about 300 AD. The "villa lobby" also point out that the format of Diocletian's Palace - that is, a walled country residence - is to be found in subsequent North African villas (known from mosaics): Diocletian's villa at Spalato was thus not an isolated phenomenon. If one makes allowance for the obvious differences of scale and elaboration, it can be seen to reflect the same current of events as that which produced a whole series of fortified villas, quasi-military in plan, which began to emerge on the provincial scene during the closing years of the third century (Boethius & Ward-Perkins 1970, 527).

Here is just such a villa, seen in a 3rd century mosaic from Carthage, and now in the Bardo Museum, Tunis:

The basic design as regards the airy piano nobile colonnade we find at Split on the sea front has been a feature of Roman villas probably since the late Republic, as seen in the view of a Campanian seaside villa from a house at Pompeii (destroyed 79 AD):

Thus the villa of Piazza Armerina in Sicily (cool as a parallel, now that few accept it as an Imperial villa, let alone as the retirement villa of Diocletian's companion-Emperor, Maximian), is of approximately the same date, and also has "interestingly" shaped rooms. Better still, the recently excavated villa at Gamzigrad in eastern Serbia, has been identified as the villa-residence of the Emperor Galerius, who took over as Augustus from Diocletian. What is more, it has palace-like structures inside (Wilkes 1986, 66ff. for a description).

Another very likely source is the palace which Diocletian as Emperor built for himself at Antioch, known to us only through a description by his contemporary Libanius:

The whole of it is an exact plan, and an unbroken wall surrounds it like a crown. From four arches which are joined to each other in the form of a rectangle, four pairs of stoas proceed ... towards each quarter of the heaven. Three of these pairs running as far as the wall, are joined to its circuit, while the fourth is shorter but is the more beautiful ... since it runs toward the palace, which begins hard by, and serves as an approach to it... (quoted from Boethius & Ward-Perkins 1970, 527).

Split has the same stoas, or covered colonnades, the same focus on the palace, and the spectacular crossing point of the four colonnades, forming a similar design to Antioch's tetrapylon.

The source of the Palace of Diocletian in the Roman military camp is quite clear - as clear as the Hellenistic and then Roman tradition of colonnaded streets, and Wilkes (1986, 60) sees the military significance of Peristyle, porch and vestibule is intelligible as a provision for military ceremony implied by the arrangement in contemporary forts. Certainly, comparisons can be made with the layout at sites such as Timgad, where the Via Principalis and the Via Praetoria adopt a similar plan (Boethius & Ward-Perkins 1970, 526).

This is fine as far as it goes - but what military ceremonial is associated with a retired Emperor? And is it known definitively that Diocletian took no part in Imperial (or ex-Imperial) ceremonial?




Press one of the other buttons, or touch the marble model.


Model of Palace (select this for a list of options)