The move away from studying Roman history almost exclusively in prosopographical terms has come to influence even those studies that are essentially prosopographical in nature. A. R. Birley's Fasti of Roman Britain (Oxford 1981) (hereafter FRB) was dominated by a prosopographical catalogue of the governors of Roman Britain, comites of the emperors, and other office holders. By contrast in The Roman Government of Britain (hereafter RGB), in addition to revising, updating, and rewriting the FRB, Birley has varied and somewhat softened the prosopographical approach with the aim of producing 'a kind of handbook or narrative history of Britain under Roman rule' (p. vii).
Most importantly, Birley has supplemented the prosopographical material with sections of narrative which provide historical and administrative background, overview, and analysis. Rather than launching straight into the career of the first governor of Britain, Aulus Plautius, as he had done in FRB, Birley now considers pre-Claudian Roman imperial attitudes to Britain (pp. 15-17) before turning to an account of Plautius (pp. 17-25) that now includes translated quotations from the literary sources dealing with the invasion (pp. 17-19), and a brief notice of Claudius' motive for invading and exploitation of its success (pp. 24-25). Short sketches introduce prosopographical sections on the governors under the Flavians (p. 57), from Nerva to Hadrian (p. 100), from Antoninus Pius and Commodus (p. 136), under Severus and Caracalla (pp. 181-83), the Governors and Legionary Legates in the third century (p. 337); other narratives treat imperial visits (e.g. Hadrian's pp. 121-22, Severus' pp. 195-203) and the Recovery of Britain (pp. 388-93). These sections of narrative are most useful when read in the context of a given period. In this sense RGB works well as a handbook for prosopographical material and brief context, whereas reading the book cover to cover does not give one a particularly satisfying experience in narrative history. For that one would have to turn to a more conventional work such as S. Frere's Britannia: a history of Roman Britain. Despite his aim, Birley shrinks from suggesting that RGB replace Frere (p. vii n. 2). Rightly so. Aside from the fact that RGB focuses on only one aspect of the Roman presence in Britain (The Government ...), the prosopographical material, which in its treatment of officials' careers is often not immediately relevant to Roman Britain, makes it difficult to keep the thread of the story in one's mind.
Other changes have helped make RGB more accessible than FRB. The prosopographical entries are numbered for ease of reference. The literary and material evidence has been translated. This supplements the sections of narrative and aids the interpretation of the sources. Birley quotes the Latin but one has to take his translations of Greek on faith (cf. the quotation of Josephus' Greek on p. 62: a mistake?). Some items regrettably slip through the net: readers will search elsewhere in vain for a translation, for example, of Mommsen's Latin on p. 458 n. 130 (quoted in FRB p. 343 n. 8). Birley has also provided a glossary of (mainly Latin) technical terms. Welcome improvements of these kinds are evidently aimed at the student and non-specialist reader, who might also have benefited from a map or two and a list of Roman emperors.
On the other hand, some material omitted from RGB might have been retained. Part III of FRB, on origins and careers, has been discarded and some of its material incorporated into what is now a shortened version of the introduction on the senatorial career under the principate. But Birley still clearly considers the introduction to FRB important enough to refer back to it for fuller discussion. Could not more of the original introduction have been retained so as to avoid some of these backward glances? The material is interesting and relevant to scholars working on the principate, not only to those with an interest in Roman Britain.
Birley has revised the prosopographical entries with care and good sense, and he demonstrates a command of the new evidence that has come to light since FRB was published. Significant among discoveries are the Vindolanda tablets, which are quoted and translated (see, for example, under Governor no. 15 and Legionary Legate no. 12). He is also able to add some new personalities to the prosopographical lists: Governors: no. 20; Legionary Legates: nos. 12, 14, 33, 36; Tribunii Laticlavii: nos. 1, 3, 9, 19; amongst the lower ranks: one procurator, two junior procurators, and a prefect of the fleet. Of the new additions the best known is the historian Cornelius Tacitus, possibly Tribunus Laticlavius '?under Vespasian, c. 77-79?' (p. 281).
Birley's revision, updating, and expansion of FRB has produced in RGB an indispensable companion to the study of Roman Britain. The book is handsomely packaged in a brightly coloured--almost lurid--dustjacket that depicts a coin celebrating the recovery of Britain under Constantius (see discussion on p. 393). But attractive dustjackets are all good and well provided the book has been made well. RGB has the same dimensions as FRB (though the type still seems too big for a book of this size), but within the book's covers FRB presented better, with its blue endpapers (alas, long gone from OUP books) and cream-coloured paper: the white paper of RGB is much too bright and the black ink much too dark. OUP's authors deserve better in this regard.
I append some specific comments:
p. 29 n. 54 In his discussion of the textual problem at Tacitus Annals 12.31.2 Birley states that the Leidensis manuscript reads cunctisque and that it 'might justify' reading cunctosque (the second Medicean manuscript has cunctaque). But the Leidensis has iunctisque, which is clear from the Mendell-Pol facsimile (1966) and accurately reported in the editions of Weiskopf (1973) and Wuilleumier (1976). In any case, since the Leidensis does not preserve a tradition independent of the second Medicean its readings have no more authority than other modern conjectures.
p. 41 For Q. Veranius' four-month ordinary consulship in 49, see now A. Tortoriello, I fasti consolari degli anni di Claudio (Rome, 2004) pp. 422-23, 585-88.
p. 43 For the consulship of C. Suetonius Paullinus (p. 43 'cos. a. inc. c.45', p. 48 'surely before 47'), see Tortoriello, pp. 416, 567-69 for a suffect consulship dated '42/44?' and a possible second suffect consulship in 66.
p. 51 The fortunes of A. Vitellius under Claudius and Nero (let alone in 69) do not suggest that 'the influence of the Vitellii declined after L.Vitellius' death in the early 50's.
p. 72 n. 42 M. Junius Silanus (PIR(2) I 839; RE s.v. Iunius 175) is unlikely to have been a 'Torquatus'. The cognomen was taken by his second son, D. Junius Silanus Torquatus (PIR(2) I 837; RE s.v. Iunius 182). See my 'Gaius and the nomenclature of military success', American Journal of Ancient History (forthcoming).
p. 81 Birley comments on Tacitus Agricola 21.2: 'as for the apparently cynical comment on the Britons not realizing that they were being enslaved, Tacitus surely meant it as favourable to Agricola'. But the sentiment of this passage is deeply cynical, and the conclusion damning: idque apud imperitos humanitas vocabatur, cum pars servitutis esset. Tacitus sees the ultimate results of Agricola's civilising measures to be grim (paulatimque discessum ad delenimenta vitiorum ...), but some of his contempt is also directed at the imperiti who choose not to share his cynicism. Tacitus does perhaps distance Agricola from the worst evils by his use of paulatim.
p. 221 Birley claims that Valerius Asiaticus was a 'prime mover' in the plot to assassinate Gaius. Although in his trial in 47 Asiaticus was labelled a praecipuus auctor of the conspiracy by Sosibius (Tacitus Annals 11.1.2), it may be seriously doubted that he had a hand in Gaius' downfall. His appearance after the event and stated wish that he had been involved suggests that he was not, but was looking to exploit the situation (cf. Josephus AJ 19.159; Dio/Xiph. 59.30. 2). Asiaticus had few reasons for acting against Gaius: he was a close friend of the family and he enjoyed considerable status. It is noteworthy too that Asiaticus' candidacy for the throne was 'blocked' by L. Annius Vinicianus, one of the ringleaders of the plot (Jos. AJ 19.252).
p. 234 Tacitus records eastern affairs in his narrative year of 47 (Annals 11. 8-10), but he narrates events that actually date between 41 and 47 (see pp. 79-81 of my article "The Date of Corbulo's Campaigns in Lower Germany" Museum Helveticum 62  76-83). This fits Birley's suggestion that C. Hosidius Geta must have been in Iberia between his service in the invasion of Britain (43) and his appointment as rex sacrorum (probably 47/48), which forbade the holder from taking up public office.
p. 255 Birley suggests that the Curtius Rufus of Tacitus Annals 11.21 might have been the 'product of a liaison between a legionary and a woman camp-follower'. But Curtius' wealthy connections (11.21.2) point to affluence on his mother's side.
I noticed some misprints: p. xiii Syme's Roman Papers is missing a date of publication, and it is referred to inconsistently throughout the book: compare p. 69 n. 31 with, e.g., pp. 107 n. 28, 128 nn. 126 and 132. -- p. 102 in the last sentence of the paragraph starting 'The Hadrianic consular...', delete the second 'of'. -- p. 198 n. 45 read: NS 2 (1994). -- p. 247 n. 74 read: (1974). -- p. 337 read: 'Britain formed part of ...'. -- p. 407 in the translation, Pan. Lat. vet. 6(7).7.4 should start at 'This fact indeed ...'. -- p. 411 in the quotations from RIC, read: mint-marks. -- p. 466 n. 2 read: ILS 216.