American Journal of Philology

The four decades between the accession of Diocletian on 20 November 284 and the abdication of Licinius on 19 September 324 witnessed profound changes in the government and administrative structure of the Roman world. These changes are naturally reflected in the legislation of the period, which survives in relative abundance—indeed, a larger proportion of the emperors’ legal decisions and enactments probably survives from these four decades than from any previous period of forty years in the history of the Roman Empire. The high rate of survival is due in part to the activity of the commissioners who compiled the Theodosian Code in the 430s and the Codex Justinianus in the sixth century: the former set out to collect general laws and edicts of the Christian Roman Empire since the conversion of Constantine, while the latter included in their compilation a large part of two Diocletianic collections of imperial rescripts, the Codex Gregorianus and the Codex Hermogenianus. But the legislation of Diocletian and his imperial colleagues and successors is also known from a wide variety of other sources—papyri, inscriptions, nonofficial collections of laws, Lactantius’ account of the decade of Christian persecution from 303 to 313 in his On the Deaths of the Persecutors, and Eusebius’ Ecclesiastical History. Moreover, Eusebius’ Life of Constantine quotes important documents from the last months of 324.

Simon Corcoran includes all this varied and historically significant material in his purview, and he provides a reliable guide to it. His monograph is a revised version of an Oxford doctoral thesis which he began in 1987 under the supervision of Fergus Millar and completed in 1992 under that of John Matthews, and he has also taken advice from Tony Honoré. His work is technically proficient. I have not noticed any careless errors of any consequence, and Corcoran has used my work on the period of Diocletian and Constantine in the spirit in [End Page 145] which I intended it to be used: he has understood the underlying principles upon which it rests, then applied them critically, sometimes to reach conclusions which not only differ from my own but are superior to them.

Corcoran begins with two introductory chapters which succinctly survey the subject and the evidence available. On interpolation in legal texts, he is cautious and sensible. He proceeds on the assumption that though “the transmitted texts are seldom identical with what the imperial chancery originally issued,” they have seldom undergone drastic alteration: “abbreviation is the most common fate of constitutions in the legal sources” (19). Yet he provides a salutary reminder and example of what can happen (16–18): whereas a rescript in the Fragmenta Vaticana denies that the temporary gift of estates subject to tax is valid (Frag. Vat. 283, si <praediorum> stipendariorum proprietatem dono de-disti, . . . donatio inrita est), the same rescript in the Codex Justinianus directly contradicts and asserts that such a gift is valid (CJ 8.54.2, si praediorum proprietatem dono dedisti, . . . donatio valet). On a priori grounds alone, it seems highly probable that something similar has happened to some texts which are known only through the Codex Justinianus—and where the alteration may not be at all obvious.

What of the date and contents of the two Diocletianic codes? Corcoran confronts the problems of dating and attribution squarely and honestly. Most of his conclusions may not be new, but his discussion has been able to proceed from the fact that an Aurelius Hermogenianus, who can hardly be anyone other than the jurist, is now securely attested as praetorian prefect between 293 and 305 with Julius Asclepiodotus as his colleague (AE 1987, 456). Corcoran argues that Gregorius published the Codex Gregorianus in 291/2 or ca. 292, after serving as magister libellorum with Diocletian from 285 to 290: he accounts for the ascription of some later constitutions to this code by postulating a second edition compiled ca. 306, possibly at the court of Maximinus or...