Modern scholarly interest in fragmentary historical texts has never ebbed since the publication of H. Peter's Historicorum Romanorum Reliquiae in 1906 and F. Jacoby's groundbreaking Die Fragmente der griechischen Historiker, the first volume of which was published in 1923 and publication of which still continues. However, in recent years there seems to have been an increased attentiveness to historical texts that have only come down to us in fragmentary form – in particular fragmented texts concerning Roman history. The most important publication lately is of course The Fragments of the Roman Historians by T. Cornell and his team of collaborators, which appeared in three volumes in 2013. This monumental and state-of-the-art work has set a new standard for the field and is a comprehensive replacement for Peter's collection from the beginning of the twentieth century. It includes the fragments of 111 Roman historians in both Greek and Latin up to the first half of the third century, thereby deliberately leaving out the fragments from the later Roman period – considered by Cornell et al. as a new phase in Roman historiography in particular because of the rise of Christian history-writing. This point of view is debatable and should be nuanced. Of course, new forms of historiography arose or became popular in the later Roman period, including Christian historiography, the chronicle and the Breviarium, but traditional historiography certainly persisted. The fourth-century historian Ammianus Marcellinus, who carried on where Tacitus had left off, is probably the best example.

It is therefore a good thing and a valuable addition to the collections of fragmentary Roman historians that Van Hoof and Van Nuffelen have made available a collection of fragments of twenty Roman historians writing in Latin from 300 to 620 ce, as well as three fragments by authors for whom it is uncertain whether their work can be qualified as historical. Many of these fragments have not been published before. Late antique historiography is one of the main research topics of Van Hoof and Van Nuffelen. They have made a huge contribution to scholarship by constructing and making available an online database of all attested late antique historiographical works by over 700 authors dated to the period 300–800 ce, the Clavis Historicorum Antiquitatis Posterioris (CHAP) (https://www.late-antique-historiography.ugent.be/database, the database has also been published in physical format: P. Van Nuffelen and L. Van Hoof [edd.], Clavis Historicorum Antiquitatis Posterioris. An Inventory of Late Antique Historiography (a.d. 300–800) [2020]).

The fragments in the volume under review have been carefully selected according to the principles: (1) That they are attested in works of later authors but do not occur in the manuscript tradition – reconstructed and/or hypothetical works such as the Enmannsche Geschichte as well as partial works (e.g. Ammianus Marcellinus) are excluded; (2) That overlap with other collections, such as The Fragments of the Roman Historians, is avoided; (3) That other genres, such as historical poetry and biography, are excluded; supposed sources used and mentioned by Nennius (ninth–eleventh century) in his History of the Britons are also excluded because it is uncertain whether these texts truly existed. In line with Jacoby's methodology, only fragments that are referred to in the source tradition have been included and texts reconstructed on the basis of modern Quellenforschung have been left out. With respect to genre, only fragments of secular history are included in the collection, and ecclesiastical history, sacred history, chronicles and, as already noted, biography are omitted. What remains then are fragments of secular prose history that can be divided into groups: fragments that are geographically limited (e.g. focusing on a city), fragments that cover a wide time span of Roman history, fragments that focus on the history of successor kingdoms, and fragments that cannot be classified because of lack of information as to their historical character.

Van Hoof and Van Nuffelen make the important observation that because the circulation of most of these texts was limited geographically and many of the texts never spread beyond the social network of the author, it is not only understandable that they were forgotten and lost, but also that claims by modern scholars that works of some of the authors, particularly Symmachus the Younger, Nicomachus Flavianus and Cassiodorus, were widely read and used and left unrecognised traces in many late antique historiographical writings are unsustainable. Another important observation that Van Hoof and Van Nuffelen make is that interest in the Roman past belonged to the social habits of the elites and has very little or nothing to do with the polemics between paganism and Christianity in the late antique period (as some historians like to argue).

Van Hoof and Van Nuffelen do not speculate as to why these histories that are only known fragmentarily through the texts of others were lost. Possibly they added nothing new to the histories that were already circulating, particularly when they were compilations of existing texts; alternatively, they may have been substandard in literary quality or only have dealt with restricted time spans or geographically limited topics. Whatever the answer, it is not very likely that, had these histories been preserved, they would have resulted in a different historical perspective; at most, they would have nuanced that perspective.

As historiography was closely connected to social status, education and reputation for the Roman state, the disintegration of the western Roman empire, social change and the decline of the traditional educational system led to the gradual demise of historiography of which the Roman state was the focal point. Attention shifted, for instance, to writing the history of the successor kingdoms and to Christian historiography.

The twenty fragmentary histories included in this volume add considerably to the scholarly perspective on late antique history writing in Latin. The following historians (in chronological order) are discussed: Carminius, an anonymous author who wrote a history of Padua, Virius Nicomachus Flavianus, Nummius Aemilianus Dexter, Protadius, Naucellius, another anonymous author who wrote a history of Rome, ps.-Hegesippus, Sulpicius Alexander, Renatus Profuturus Frigeridus, Favius, Consentius, Ablabius, Symmachus the Younger, Maximian of Ravenna, Marcellinus Comes, Cassiodorus, Roterius, Secundus of Trent and Maximus of Zaragossa. The three authors for whom it is doubtful whether they wrote histories are Bruttius, Latinus Alcimus Alethius Rhetor and Tyconius.

The format of the entries is similar. First there is an introduction presenting the writer and his dates, the (possible) title and the character of his work, and an engagement with the scholarly literature on the fragment and its author. Then follows the Latin text of the fragment and an English translation, accompanied by a commentary. Depending on the fragment, the commentary engages with philological and historical concerns, issues of interpretation and other matters important for a better understanding of the text.

Within the context of this review it is impossible to review all fragments individually. Let me therefore present a few examples. Three letters by Symmachus – all probably written in 396 – testify that Protadius, a member of the aristocracy of Gaul and praefectus urbi in 401, wrote a history of Gaul. In their commentaries of the relevant passages in the letters Van Hoof and Van Nuffelen conclude convincingly that Protadius’ history concerned the early history of Gaul, most likely the Roman conquest and possibly pre-Roman Gaul, and that it was a compilation of earlier histories by Caesar, Livy and Pliny the Elder, who are explicitly referred to in one of the letters. Beyond Symmachus, Protadius’ history is not mentioned, which makes it very likely that the work only circulated within Protadius’ own social circle.

The history of Sulpicius Alexander is only known from Gregory of Tours, who refers to it seven times and used information provided by Sulpicius in Books 3 and 4 of his work on the leaders of the Franks at the end of the fourth century – the fragments cover the years 388–393. About the character of Sulpicius’ history little can be said, nor where in time his work started or when it ended. It has been often assumed that Sulpicius continued from where Ammianus Marcellinus finished his Res Gestae, i.e. from 378, and that his history was of similar character to that of Ammianus, but Van Hoof and Van Nuffelen rightly call this into question. Instead, they suggest that Sulpicius’ history was a work along the lines of Zosimus: the first books (Books 1 and 2 in the case of Sulpicius) rapidly narrate the rise of Rome, and the author then slows down in Books 3 and 4 (and possibly later books) to treat the history of the fourth century. This may be so, but it is no more than an educated guess.

Of most authors in this volume no other works are known apart from the fragments, but this is not the case for Cassiodorus, who held various offices under Theodoric the Great and also served under his successors. Cassiodorus is best known for his Variae. Based on a thorough analysis of references made by Cassiodorus himself in the Variae and by Jordanes, Van Hoof and Van Nuffelen conclude that Cassiodorus created a work about the origin and the deeds of the Goths, and in particular their leaders, based on earlier source material. The work was not only historical but also ethnographical in character, and probably bore the title History of the Goths. The Getica by Jordanes follows Cassiodorus’ narrative, although it is hard to establish to what extent the Getica is a reflection of Cassiodorus’ History of the Goths or to what degree Jordanes intervened in Cassiodorus’ text. Van Hoof and Van Nuffelen argue that the work was written at the request of Theodoric and was also dedicated to him, implying that it was written sometime between 515 and 526. However, they argue against a political overinterpretation of the history or its reading as a work legitimising the Amal dynasty. According to Van Hoof and Van Nuffelen the History of the Goths had a limited circulation, and after Jordanes there are no traces of use and knowledge of the History in any work, including, despite the arguments of some scholars, the Anonymus Valesianus.

This is an important contribution to the study of late Roman historiography and literary activity in late antiquity in general. Corresponding to Jacoby's guidance the approach by Van Hoof and Van Nuffelen is careful and down to earth: they present what is known and do not hesitate to emphasise what is not known, and they caution against speculation and overinterpretation of the fragments. Given that Greek was the dominant language in late antiquity we are looking forward to the collection of Greek Fragmentary Chronicles after Eusebius that Van Hoof and Van Nuffelen are now preparing.